computer-implemented method for virtual adjustment of glasses

专利摘要:
The present invention relates to a process, a device and a computer program for the virtual adjustment of an eyeglass frame. A process for the virtual adjustment of glasses is provided, as well as a corresponding computer program and a computing mechanism. In this case, first measuring points are defined in a 3D model of a person's head, and a model of an eyeglass frame is adjusted based on the first measuring points. According to the invention, the definition of the first measurement points comprises a definition of second measurement points in a parametric head model, an adjustment of the parametric head model to the 3D model of the person's head and a determination of the first measurement points based on second measurement points and adjustment. In this way, the second measuring points have to be defined only once in the parametric head model, so that the first measuring points can be defined for a plurality of different 3D models with different heads.
公开号:BR112020000106B1
申请号:R112020000106-6
申请日:2018-07-03
公开日:2021-01-05
发明作者:Oliver Schwarz；Ivo Ihrke
申请人:Carl Zeiss Ag；Carl Zeiss Vision International Gmbh；
IPC主号:

专利说明:

[0001] This application relates to methods, devices and computer programs for the virtual adjustment of eyeglass frames.
[0002] An eyeglass frame means, in this case, a frame or support in accordance with DIN ESO 77998: 2006-01 and DIN ESO 8624: 2015-12, through which / from which glasses can be carried on the head. The term used here includes, in particular, rimless spectacle frames. Eyeglass frames are also colloquially referred to as eyeglass frames. A virtual placement of an eyeglass frame designates, in the context of this application, an adjustment of a model of an eyeglass frame to a model of a head in a computing mechanism, usually associated with a graphical representation of the adjustment of the eyeglass frame on a person's head on a display, for example, a computer screen.
[0003] A virtual placement of an eyeglass frame on a head is known, for example, from US 2003/0123026 A1 or from US 2002/015530 A1. In these publications, the virtual placement of the eyeglass frame serves mainly to facilitate a user in selecting between different eyeglass frames, by displaying a graphic representation of the user's head, together with the eyeglass frame.
[0004] US 9,286,715 B2 discloses a method for virtual eyewear testing. In this case, several points are defined both on a spectacle frame and on a head. Positioning the glasses frame on the head is done by combining selected points on the glasses frame with selected points on the head. A change of position occurs by changing the selected points. This allows a positioning with an accuracy that is sufficient for the purpose of document US 9,286,715 B2 to provide a proof to obtain a visual impression. Similarly, US 2005/162419 A discloses a virtual placement of an eyeglass frame with the help of characteristic points. In this publication, a frame is first scaled and then positioned in different directions. Finally, eyeglass frame rods are rotated around two spatial axes.
[0005] A video demonstration of the software "Vacker" is available at „https://www.volumental.com/face-scanning/", of March 5, 2017 by the company Volumental, in which a head with some glasses placed and glasses parameters can be modified by means of sliding bars, for example, the adjustment of the glasses on the back of the nose or also other parameters, such as the angle of the frame rims, in addition, a color of the glasses frame or a color of the glasses frame hinges can be selected.The selected parameters are then emitted In this video, different parameters of a parametric model of a glasses frame are also adjusted.
[0006] Another system for the virtual adjustment of glasses is known from US 2015/0055085 A1. Here, an automatic adjustment of the glasses is made by adjusting the size and adjustment of the glasses on a person's head. In addition, the shape, style and color of the glasses can be selected.
[0007] A method and device for projecting bespoke glasses, that is, glasses fitted to a person's head, are known from DE 10 2016 824 A1. In this method, image data from the head is captured in two or three dimensions, sample glasses are selected and sample glasses are represented according to the projection parameters of the sample glasses. Projection parameters are determined based on image data from the head.
[0008] US 2015/0277155 A1 discloses an individualization of the frame of an eyeglass frame, in which distances are measured on a person's face and in which an eyeglass frame is created using 3D printing based on the distances measures.
[0009] US 2013/0088490 A1 discloses an iterative method for adjusting an eyeglass frame, with the eyeglass frame being positioned in small steps and the adjustment occurs based on a collision detection, with the which is checked if the glasses frame overlaps the person's head.
[0010] The document US 8733936 B1 discloses a method and a system for fitting glasses to a person's head with the help of image captures of the head.
[0011] The document WO2014 / 037915 A1 also discloses a method and system for adjusting glasses to a person's head, with characteristic points being identified in an image of a person's face, to which the glasses must be adjusted.
[0012] Other methods and systems for adjusting glasses are known from US 9703123 B2 or US2016 / 062152 A1 on a person's head, for example, based on anatomical data, which are entered into the system by a user .
[0013] The document US 2015/0293382 A1 discloses a parameter determination for a virtual sample of glasses by means of capturing a person with the exemplary frame placed. The parameters that have been determined using this example frame are modified accordingly to a virtual sample of a virtual frame. Since the person already wears an eyeglass frame during capture, no three-dimensional model of the head is used without the eyeglass frame.
[0014] Niswar, Kahn and Farbiz describe in the article "Virtual Try-On of Eyeglasses using 3D-Model of the Head", Institute for Infocomm Research, December 2011, issue date: 10.1145 / 2087756.2087838, a method for the virtual sample with glasses. This is based on four reference points, with two points on the nose and two points on the ears. In this case, a 3D head model is adjusted by deforming a generic head model based on some feature points.
[0015] US 2016/0327811 A1 describes a method that is based on a virtual model of a frame. This is adjusted to a head through deformation. For the adjustment of the glasses frame, adjustment criteria can be performed, for example, a maximization of a platelet contact surface in relation to a person's nose, a maximization of a contact surface of spectacle rods, a centralization a frame edge of the glasses frame in relation to the eyes, an orientation of the glasses frame or a minimization of the contact surface of the frame edge in relation to the person's cheekbones and in relation to the person's eyebrows.
[0016] As a possible extension of these criteria, an establishment of target values is mentioned. Target values of this type may refer, for example, to a distance between both spectacle arms of the spectacle frame, an inclination towards the front of the frame, a distance of the frame's platelets from one another, a distance of one eye from the rim of the frame, a distance from the rim of the eyebrow and cheekbones, an inclination to the front of the spectacle frame or an angle of the rim of the spectacle frame frame. These parameters and target values are included in a cost function and through a conventional optimization method, for example, a Levenberg-Marquardt algorithm, an optimization is performed. The frame can then be further deformed.
[0017] In this method, it is problematic that, in such an optimization method, a global optimum is not necessarily achieved, since, with optimization methods, such as the Levenberg-Marquardt algorithm, only one local minimum cost function. In the case of surface oscillations of 3D models used for eyeglass or head frames, it may happen that the optimization of a surface oscillation of this "estagne" type is far from ideal and, therefore, no optimal adjustment is achieved.
[0018] In addition, an optimization using such an optimization method is very computationally intensive, in the case of using many parameters. This makes it difficult to use parametric frame models, in which a greater number of parameters have to be optimized.
[0019] In a series of publications referred to above, for example, US 9,286,715 B2, US 2005/162419 A1 or US 2016/0327811 A1, points are marked on the 3D model of the head and then used for adjustment of the glasses frame model, for example, points on the nose of the head. These points are essentially taken for granted. US 9,286,715 B2 generally refers to image processing to obtain these points, and US 2016/0327811 A1 also does not give any exact indications and makes reference to a computerized determination.
[0020] US 2005/0162419 A1 discloses a method for establishing points on a head model, in which a 2D image of a person is captured and in which 2D image points are marked. These points are then transferred to a standard head model. Based on these points, an eyeglass frame is then adjusted.
[0021] Here the user also needs to mark the corresponding points on the person's face, which takes time.
[0022] WO 2016/164859 A1 discloses two different possibilities for obtaining a 3D model of a person's head. In a first approach, a generic parametric head model, for example, based on anatomical parameters, which are introduced by a user, are adjusted to the person. This adjustment can also happen when certain characteristics of the parametric model are harmonized with images of the person. In another approach, a parametric model is generated again based on the person's anatomical data. The parametric model can, in both cases, be particularly detailed in fields that are relevant for the positioning of eyeglass frames. For glasses adjustment, relevant measuring points can also be defined, in this case, otherwise, based on image captures. Additional image captures are also required here.
[0023] Based on US 2005/0162419 A1, an objective of the present invention is to provide a method for adjusting glasses, as well as a corresponding computer program and a device, in which an establishment of points on a head model 3D of a person, at least for a part of the points, can occur in an automated way, without a user having to mark the points individually on a face or on the model of a person's head, to which glasses must be adjusted, and without the need for an image capture analysis, as in WO2016 / 164859A1.
[0024] According to the invention, a method implemented by computer or a method performed by a computer for virtual adjustment of glasses is made available, in which first measuring points are defined in a 3D model of a person's head and in which model of a spectacle frame is fitted to the 3D model of the person's head based on the first measurement points. The method is characterized by the fact that the definition of the first measurement points comprises:
[0025] adjust a parametric head model to the 3D model of the person's head, and
[0026] determine the first measurement points based on second measurement points defined in the parametric head model and the adjustment of the parametric head model to the 3D model of the head.
[0027] By measuring points, in this case, general points are understood in a model, which can be used for the subsequent adjustment of glasses, particularly for the measurement of sizes, such as distances on the head.
[0028] Through this method, the second measurement points in the parametric head model only need to be defined once. This parametric head model with the second measurement points defined in it can then be used for 3D models of the heads of different people, to define the first measurement points in these 3D models. Unlike the WO 2016/164859 A1 approach, therefore, no images of the person or other information are considered, and two models are used, namely the parametric head model and the 3D model of the person's head, whereas, in WO 2016/164859 A1, only one model is used.
[0029] The unique definition referred to above the second measurement points in the parametric head model can occur in the context of the method above or also separately and / or previously, for example, in another computer. Correspondingly, the above method can comprise a step of defining the second measurement points in a parametric head model, or the second measurement points can be defined in advance and made available, for example, when a method comprising the step is made available separately the definition of the second measurement points in a parametric head model. The second measurement points can then be used for a plurality of different 3D models depending on the person, without the points having, for example, to be manually defined for each 3D model.
[0030] The concepts used in the method referred to above and in the method described below are further explained.
[0031] The setting is "virtual", as the procedure takes place in a computing mechanism, for example, in a personal computer (Personal Computer) (PC) and the real glasses frame is not placed over the real head.
[0032] A model, particularly a 3D model, means a three-dimensional representation of real objects, which are present as a data set in a storage unit, for example, in a computer storage or in a data carrier. This three-dimensional representation can be, for example, a 3D mesh (in English "3D mesh"), which consists of a set of 3D points, also called vertices, and connections between points, also called edges. This connection, in the simplest case, forms a triangular network (in English triangle mesh). In the case of such a representation as a 3D network, only the surface of an object is described, not the volume. The network does not necessarily need to be closed. If, for example, the head is described in the form of a net, it will appear as a mask. You will be able to find out more about these 3D models in Rau J-Y, Yeh P-C. "„ A Semi-Automatic Image-Based Close Range 3D Modeling Pipeline Using a Multi-Camera Configuration ". Sensors (Basel, Switzerland). 2012; 12 (8): 11271 to 11293, issue date: 10.3390 / s120811271; particularly page 11289, image "Figure 16").
[0033] Another possibility of representing a 3D model is a grid of columns, which represents a volumetric representation. The space is divided into small cubes or cuboids, called vóxis. For each voxel, in the simplest case, the presence or absence of the object to be displayed is stored as a binary value (1 or 0). With a border length of 1 mm and a volume of 300 mm x 300 mm x 300 mm, which represents a typical volume for a head, a total of 27 million of these are obtained. Such gridlines are described, for example, in M. Nieβner, M. Zollhofer, S. Izadi, and M. Stamminger. "Real-time 3D reconstruction at scale using voxel hashing". ACM Trans. Graph. 32, 6, Article 169 (November 2013), DOI: https://doi.org/10.1145/2508363.2508374.
[0034] The 3D model of the head and / or the 3D model of the glasses frame can, in particular, be a 3D model with texture. A 3D model with texture is a 3D model that also contains the color information of the surface points of the real object. Through the use of a 3D model with texture, a realistic color representation of the head and glasses frame is possible.
[0035] In this case, the color information can be contained directly at the vertices as an attribute, for example, as an RGB color value (red-green-blue), or a pair of texture coordinates is attached to each vertex as an attribute. The vertex designates, as mentioned above, a point on the 3D model. An attribute then designates, in general, a characteristic, indicator or similar, which is assigned to an object, in this case, to its respective vertex (see also Wikipedia article "Attribut (Objekt)", of 5 July 2017) . These coordinates must be understood as image coordinates (pixel positions) in an additional texture image. The texture, for example, of the triangles of the triangular lattice mentioned above is generated by interpolation from the pixels of the texture image.
[0036] A parametric model is a 3D model, which has one or more changeable parameters. By changing the parameter or parameters, the geometry of the object described is then changed using the 3D model, in this case, the glasses frame, for example, with respect to size or shape. Examples of such parameters are, for example, a width of the bridge or a length of the spectacle frame rods, or also a shape of a spectacle frame edge. The type and number of these parameters depends on the glasses frame represented by the parametric frame model. In particular, ranges of values for the parameters can be established by an eyeglass frame manufacturer, which then describe, correspondingly, eyeglass frames that can be manufactured. A free frame parameter means a parameter of the parametric frame model that has not yet been established in the method, that is, it still needs to be adjusted and determined.
[0037] Adjustment requirements are specifications of how the glasses frame should be positioned in relation to areas or points of the head, such as glasses, pupils, eyebrows or nose. These specific adjustment requirements for the parametric frame model are particularly taken advantage of to ensure an aesthetic impression desired by the eyeglass frame manufacturer. Specific adjustment requirements can be made available in conjunction with the parametric frame model of a respective manufacturer in electronic format, for example, as corresponding files.
[0038] The anatomical adjustment refers, on the other hand, to an adjustment that should guarantee a correct comfortable fit of the glasses frame on the head. For this purpose, criteria are used, which are not specific to the respective glasses frame, but generally apply to a plurality of glasses frames, such as correct adjustment of the glasses' stems over the ears or a correct adjustment of the platelets of the glasses. Anatomical adjustment may also comprise ensuring minimum distances from areas of the head, for example, ensuring a minimum distance from the edges of the glasses frame of the cheekbones and / or a portion of the eyebrows from the head and / or a minimum distance from the eyebrows. eyelashes. Another example of an anatomical adjustment is a regulation of a nominal distance or a nominal range for the distance between lens and eye, that is, the distance to the apex of the cornea (Hornhaut-Scheitel-Abstand) (HSA). The distance to the apex of the cornea is the distance between the front surface of the cornea of the eye and the surface facing the eye of the spectacle lens. For example, during anatomical adjustment, it is ensured that a nominal corneal apex distance of 12 mm or a nominal corneal apex distance between 12 mm and 17 mm is maintained. The reason for this is that the lens of the glasses should not be placed close to the eye to avoid impact of the lashes and to avoid condensation on the lens (perspiration). Still, some optometrists would like to avoid a deviation of the distance to the corneal apex from a distance to the corneal apex predefined in a phorometer used to measure spherical-cylindrical refraction. Since a greater distance to the corneal apex changes the optical effect towards positive diopter values, a distance to the largest corneal apex may be preferable in the case of hyperopia, that is, when so-called positive lenses are required. Therefore, a distance to the apex of the cornea based on the result of the refraction measurement can be advantageously used.
[0039] The adjustment requirements are present, in this case, preferably in text format, for example, as .xmal or JSON files, which facilitates processing.
[0040] "Person" means, in the context of this application, the person to whose head the spectacle frame must eventually be fitted. "User" means a person who makes and operates the device or method for adjusting the frame. This can be the person himself, but it can also be someone else, for example, an optometrist.
[0041] Suitable parametric head models are described, for example, in A. Brunton, A. Salazar, T. Bolkart, S. Wuhrer, "Review of Statistical Shape Spaces for 3D Data with Comparative Analysis for Human Faces", Computer Vision and Image Understanding, 128: 1-17, 2014 or in J. Booth, A. Roussos, S. Zafeiriou, A. Ponniahy and D. Dunaway, "A 3D Morphable Model Learnt from 10,000 Faces", 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, 2016, pages 5543 to 5552. Issue date: 10.1109 / CVPR.2016.598.
[0042] Preferably, the definition of second measurement points is made in the parametric head model through the definition of second measurement points in a standard head of the parametric head model or the second measurement points are previously defined in a standard head of the parametric head model. A standard head is, in this case, a parametric head model head, in which parameters of the parametric head model assume predefined values.
[0043] In head models based on a principal component analysis, the average value of the data underlying the principal components may be the standard model.
[0044] In this way, a defined starting point for the method can be made available.
[0045] The characteristics defined in the standard head can then be transferred to the adjusted parametric head model, corresponding to the adjustment, that is, corresponding to the change of the standard head to the adjusted parametric head model, with the seconds Measurement points are also changed, so that they are located in corresponding locations on the adjusted head model. Thus, the second measuring points can be easily transferred to the adjusted head model. The rationale for such a transfer of points is explained in section 4.1 of the reference cited above J. Booth et al.
[0046] The second measurement points transferred to the second head model can then be used directly as the first measurement points, if the adjustment of the head model is sufficiently accurate. What is sufficiently accurate depends on the desired accuracy for the following glasses setting and the accuracy of the measurement system, with which the 3D model of the head was created. Thus, for example, in the case of an accuracy of the measurement system of 0.2 mm and a desired accuracy of the points of 0.5 mm with the help of the adjustment of the parametric head model, an accuracy of also 0, 5 mm. If the measurement system has a much worse accuracy, then it depends on the distribution of the errors: in the case of a Gaussian distribution error, a smoothing is obtained - the accuracy, in the sense of the maximum deviation, is generally improved through the adjustment. Alternatively, determining the second measurement points may include projecting the transferred measurement points onto the 3D model of the head. Using such a projection, the first measuring points can then be determined in a simple way. For the projection of, for example, the respective first measurement points, a section point of a normal vector can be used on the parametric head model adjusted through the respective second measurement point transferred.
[0047] The adjustment can be carried out with the help of conventional adjustment algorithms (in English, called fitten) (see the Wikipedia article "Ausgleichsrechnung", 22 May 2017).
[0048] The method can comprise a combination of several first measurement points with a characteristic, which characterizes a field of the 3D model of the head. With the help of such features, they can then be effectively adjusted to eyeglass frames, when the characteristics for the adjustment characterize relevant fields of the 3D model. Examples of features include, for example, nose flaps, a curvature of the forehead or a base of the ears.
[0049] The combination may, in this case, comprise an adjustment of a geometric primitive or function to the various measurement points. Examples of such geometric primitives are planes, circular segments, spherical segments or cylinders. Examples of functions include spline functions. In this way, several measurement points with few parameters can be described (for example, receiving point and normal vector in the case of a plane, radius and midpoint in the case of the forehead curvature), which facilitates processing, since they are less data is present. The frame model may comprise a parametric frame model.
[0050] With the nose flaps as a feature, for example, a bridge width of the parametric frame model can be easily adjusted, for example, the platelets of the parametric frame model are covered with the nose flaps. . A stem length of the parametric frame model can be determined with a base point of the ears as a measurement point or characteristic. The pantoscopic angle of the eyeglass frame according to DIN EN ISO 13666: 2012 5.18, that is, an inclination of the rim of the frame, can be determined by determining a distance between the rim of the frame and a characteristic, which describes the cheeks of the person's 3D model. In this way, different parameters of a parametric frame model can be easily adjusted.
[0051] The method may comprise a calculation of other measurement points for the 3D model of the head based on first measurement points or on the characteristics explained above. Examples of such other calculated measuring points include an ear support point for first measuring points in the cheek area and / or first measuring points on the ear. The ear support point is, in this case, a point on which an eyeglass stem rests during the adjustment of the eyeglass frame. In other modalities, the ear support point can be one of the first measurement points, which are determined based on the second measurement points.
[0052] In such a method, another measurement point, such as the ear support point, can be calculated even in the case of hair covering. In this case, the other calculated measuring point need not necessarily be on the 3D model, but it can also be located away from it.
[0053] For the calculation of another measurement point, a predefined geometric realization between the first measurement points and another measurement point of this type can be used. In this case, a geometric relationship indicates how the other measuring point is in relation to the first measuring points. As an example of this, the calculation of another measurement point may comprise a linear combination of a number of first measurement points. As another example, the calculation of another measurement point may comprise an extrapolation based on the first measurement points. In this case, a curve or surface model, for example, a curve or surface with one or more free parameters, can be calculated based on the quantity of the first measuring points by means of interpolation or approximation by means of an error function. , for example, due to the fact that the free parameters are determined by an adjustment procedure (English: "fitten").
[0054] For example, a polynominal curve can be adjusted to the first measurement points that are located on the cheek. During extrapolation, this polynominal curve is evaluated in the zone that is located towards the ear and, thus, another calculated measurement point is determined.
[0055] In addition, a computer program with programming code is also available which, when executed on a processor, performs one of the methods described above. Finally, a corresponding device is provided with a processor with a memory, which stores the computer program, and with a processor for executing the computer program.
[0056] A computer program is also made available, comprising commands that, in case the program is executed by a computer, make it perform the method as described above.
[0057] A computer program is also available, comprising commands that, in case the program is executed by a computer, cause it to perform the following steps:
[0058] define first measurement points on a 3D model of a person's head, with measurement points being points on a model, which can be used for a subsequent adjustment of glasses, and
[0059] fitting a model of an eyeglass frame to the 3D model of the head based on the first measuring points, characterized by the fact that the definition of the first measuring points comprises:
[0060] fit a parametric head model to the 3D model of the person's head, and
[0061] determine the first measurement points based on second measurement points defined in the parametric head model and the adjustment of the parametric head model to the 3D model of the head.
[0062] In addition, other of the method steps explained above can be performed.
[0063] A computer-readable, in particular, tangible storage unit is also available, comprising commands that, in the case of execution by a computer, make it perform the method as described above. Examples of storage units include optical storage units, such as CDs or DVDs, magnetic storage units, such as hard drives or solid state memories, such as flash memories or read-only memories (ROMs).
[0064] A computer-readable storage unit, in particular tangible, is also made available, comprising commands that, in the case of execution by a computer, make it perform the following steps:
[0065] defining first measuring points on a 3D model of a person's head, with measuring points being points on a model, which can serve for a subsequent adjustment of glasses, and
[0066] fitting a model of an eyeglass frame to the 3D model of the head based on the first measuring points, characterized by the fact that the definition of the first measuring points comprises:
[0067] fit a parametric head model to the 3D model of the person's head, and
[0068] determine the first measurement points based on second measurement points defined in the parametric head model and the adjustment of the parametric head model to the 3D model of the head.
[0069] In addition, other steps of the method explained above can be performed.
[0070] A computer-readable data medium, in particular tangible, is also provided, in which the computer program is stored, as described above.
[0071] In addition, a data carrier signal (for example, over a network, such as the Internet) is provided, which transfers the computer program, as described above.
[0072] A device is also made available for data processing and / or for adjusting glasses, comprising means for performing the method, as described above.
[0073] A device is also made available for data processing and / or glasses frame, comprising:
[0074] means for defining first measurement points on a 3D model of a person's head, with measurement points being points on a model, which can serve for a subsequent adjustment of glasses, and
[0075] means for fitting a model for an eyeglass frame (120) to the 3D model of the head based on the first measuring points, characterized by the fact that the means for defining the first measuring points comprises:
[0076] means for fitting a parametric head model to the 3D model of the person's head, and
[0077] means to determine the first measurement points based on second measurement models defined in the parametric head model and the adjustment of the parametric head model to the 3D model of the head.
[0078] For that, optionally, other means can also be made available to carry out other steps of the method described above.
[0079] A device is also made available for data processing and / or for adjusting glasses, comprising a processor, which is configured to carry out the method, as described above.
[0080] The computer programs and devices described above can have the same properties described for the methods.
[0081] In addition, a method for making an eyeglass frame is made available, comprising:
[0082] perform the method, as described above,
[0083] virtual adjustment of an eyeglass frame to the 3D model of the head using the first measuring points, and
[0084] manufacture the adjusted glasses frame.
[0085] The 3D model made available as described above with the method with the first measuring points is therefore used first for the virtual adjustment of an eyeglass frame. The virtual adjustment of the spectacle frame can occur autonomously, as described in the state of the art explained at the beginning. The adjusted eyeglass frame can then virtually be manufactured as a real eyeglass frame, as in the prior art mentioned above. Manufacturing can be done, for example, using an additive method, such as 3D printing, see the Wikipedia article "Generatives Fertigungsverfahren", of 25.06.2018 for an overview.
[0086] The invention is explained in more detail below on the basis of preferred embodiment examples with reference to the accompanying drawings. Show:
[0087] Figure 1 a device for the virtual adjustment of glasses, according to an example of modality,
[0088] Figure 2 an example for an implementation of a camera mechanism from Figure 1,
[0089] Figure 3 a flowchart, which provides an overview of a method for setting the frame, according to an example of modality,
[0090] Figure 4 a flowchart of a method, according to the example of modality, which can be applied in the method of Figure 3,
[0091] Figure 5 a flowchart of a method, according to the example of modality, which can be applied in the context of the method of Figure 3,
[0092] Figure 6 a view to illustrate characteristics of a head, which can be referred to,
[0093] Figure 7 a detailed implementation of method step 40 in Figure 4 or step 54 in Figure 5,
[0094] Figure 8 a diagram to explain auxiliary characteristics,
[0095] Figure 9 schematic views of a head to illustrate an adjustment,
[0096] Figure 10 other schematic views of a head to illustrate an adjustment based on adjustment requirements,
[0097] Figure 11 a flowchart of a method, according to the example of modality, which can be applied in the context of the method of Figure 3,
[0098] Figure 12 a flow chart of a detailed implementation of the method of Figure 11,
[0099] Figures 13A to 13D and 14 representations to illustrate head models,
[0100] Figure 15 a diagram to illustrate a partial step of adjusting a pair of glasses in the method of Figure 12, and
[0101] Figure 16 is a view of a frame model to illustrate a width of the bridge.
[0102] Figure 1 shows an example of the modality of a device for the virtual adjustment of the frame, according to an example of modality. The device of Figure 1 comprises a computing mechanism 11, which has a processor 12, as well as a memory 13. Memory 13 is for storing data and comprises, in the embodiment example of Figure 1, a memory with random access ( Random Access Memory, RAM), a read-only memory (Read Only Memory, ROM), as well as one or more mass storage units (hard disk, Solid State Disc, optical disc drive, etc.). A program is stored in memory 13, whereby, when it is executed in processor 12, a method is carried out, as already described above or as to be explained in more detail below for a virtual frame adjustment.
[0103] The device in Figure 1 has a display 16, on which a person's head is displayed, together with an eyeglass frame, when the computer program is run on processor 12. User inputs can be made through one or more input devices 17, for example, keyboard and mouse. Also or alternatively, the display 16 can be a touchscreen (Touchscreen), to be able to make entries.
[0104] The device of Figure 1 further comprises an interface 14 with a network 18 through which data can be received. In particular, parametric eyeglass frame models and assigned fitting requirements from eyeglass manufacturers can be received here. In some examples of modality, through the interface 14 data can also be sent to another computing mechanism, to perform, for example, part of the calculation necessary for the adjustment of glasses. To create a 3D model of a person's head, to which the glasses must be adjusted, the device in Figure 1 optionally comprises a camera mechanism 15, through which several images of the person can be captured from different directions and the 3D model can be determined. Information on determining such 3D models based on image captures can be found, for example, in H. Hirschmüller "Stereo Processing by Semiglobal Matching and Mutual Information" in IEEE Transactions on Pattern Analysis and Machine Intelligence, volume 30, no. 2, pages 328 to 341, February 2008. Issue date: 10.1109 / TPAMI.2007.1166).
[0105] Figure 2 shows an embodiment of the camera mechanism 15 of Figure 1. In the example embodiment of Figure 2, a semicircular array of cameras 110 is attached to a column 19. A person can then position himself so that a head 111 of the person, as shown in Figure 2, is positioned in the semicircular arrangement 110, and can be captured from different directions. From there, a 3D model of the head 111 can then be created. From the image captures a texture is also formed, that is, information in relation to colors (as explained above) of the model. Such a device can also be used for centralization measurements, as described in EU patent application 17 153 556.0.
[0106] Figure 3 shows a flow chart of a complete method for the virtual adjustment of glasses, according to an example of modality. The present application refers, in particular, to partial stages of this method.
[0107] The method starts at step 30. In step 31, a 3D model of the head, including metadata of the head model, is loaded from memory. The 3D model can be created, as explained above, with reference to Figures 1 and 2, with the help of image captures or it can be a 3D model already present, for example, from a previous eyeglass fit for a certain person.
[0108] The metadata of the head model is data that contains information about the characteristics of the 3D model, but not the model itself. In particular, metadata can provide additional information to the 3D model of the head and / or contain certain points, curves or zones in the 3D model of the head. Further details on the use of such metadata are also found in EU patent application 17 173 929.5.
[0109] In step 32 a base model of an eyeglass frame is selected, which is described through a parametric frame model. The parametric frame model has free parameters, that is, to be determined. Examples of such free parameters have already been mentioned above in the description of the parametric frame model, namely the width of the bridge or the length of the glasses frame rods, or also a shape of a frame frame edge.
[0110] In step 312, at least some of the parameters are then calculated based on an adjustment requirement assigned to the frame model, as described above or as explained in more detail below. Other parameters are determined based on an anatomical adjustment, as also explained above.
[0111] In steps 33 to 310, a virtual placement of the glasses occurs with a supplementary anatomical adjustment. For this, in step 33 an approximate positioning based on a placement point and a support of the back of the nose runs, as already described in patent application EU 17 173 929.5. In steps 34 and 35 there is a folding of the glasses rods in relation to the ears of the head and a positioning of the rods, and a rotation around the x-axis of the glasses can occur. The x axis corresponds, in this case, to a direction that connects the eyes of the head, the z direction essentially corresponds to the direction of the rods and the y direction is perpendicular to it. In step 36, the contact surfaces of the glasses are optimized by means of an exact positioning in the xy plane. In addition, parameters not yet established can be adjusted here in step 312. The steps 34 to 36 correspond, in this case, to the steps correspondingly described in the patent application EU 17 173 929.5. In this setting, the parametric glasses model can, in particular, be deformed and positioned, after the parameters have been determined in step 312.
[0112] In step 37 there is then a rendering of the frame and head, that is, a corresponding representation in the display 16 of Figure 1. This rendering is also already described in patent application EU 17 173 929.5. Rendering, also called rendering or synthesis of images, in this case means the creation of an image (for example, for display on a computer screen) based on raw data, in this case, from the respective models.
[0113] A user interaction with the model then follows in step 38, which has several consequences, as represented in step 39. Thus, a simple navigation can occur, for example, to view the head from another direction . In this case, a new rendering occurs in step 37.
[0114] Through interaction in step 39, the frame rotation around the x axis can also be adjusted manually. In this case, the method returns to step 35 to determine, for example, the rods corresponding to the new frame position.
[0115] In addition, through user interaction with the model, the position of the eyeglass frame can also be adjusted to the back of the head model's nose by a device user. This essentially alters the position established in step 33 of the spectacle frame. Therefore, in this case, the method returns to step 33.
[0116] These types of interaction described so far, particularly navigation, for example, to change the viewing angle, adjust the rotation and adjust the position of the glasses arranged on the back of the nose, are also explained in detail in the order US patent 17 173 929.5.
[0117] In addition, in the interaction, one of the frame parameters of the parametric frame model can also be set by a user. For example, a user can here change the parameter determination that took place through automatic calculation in step 312. In this case, this reduces the number of frame parameters in step 310, and the method is continued in step 36. If after the interaction the Finally, if the user is satisfied with the adjustment, the method is completed in step 311. In this case, a final control may still occur. In the final control, the user (for example, an optician) checks the availability data. In this case, the order data, as well as correspondingly graphic representations, are displayed for it on an overview screen. The representations show the parameters of the glasses and / or head frame determined in the context of the method, such as the width of the bridge and the angle of the nose flaps, etc., and also the parameters of the requested frame, possibly also indications of deviations of a shape. ideal, which is pre-defined, for example, by the adjustment requirements. The determination of parameters of this type will be further explained later. The calculated parameters can then be transmitted to an ordering system from the respective manufacturer, to order a real glasses frame with the corresponding parameters.
[0118] Next, individual aspects of the method in Figure 3 are explained in more detail with reference to Figures 4 to 15.
[0119] Figure 4 shows a flow chart of a method according to an example of modality. Figure 4 shows a division of the spectacle frame into an adjustment based on adjustment requirements assigned to the respective parametric frame model, followed by an adjustment to a head anatomy.
[0120] In the method of Figure 4, in step 40 there is an adjustment of the parametric frame model to a 3D model of the person's head based on adjustment requirements, which are pre-defined specifically for the glasses frame by the manufacturer of the glasses frame of the respective glasses frame. These adjustment requirements may refer to aesthetic indications, as will also be explained in more detail below. Implementation examples for this step will be explained in more detail later. Step 40 can be performed, for example, in the context of step 312 in Figure 3.
[0121] By setting in step 40, a first part of the parameters of the parametric frame model can be established.
[0122] In step 41, a general adjustment is made to the anatomy of the person's head, that is, the adjustment in step 41 occurs independently of specific requirements. This adjustment can occur as described in the state of the art mentioned at the beginning, and it can also occur in step 312 or possibly also in the adjustment in steps 34 and 35. The anatomical glasses adjustment can then also happen directly based on the metadata of the head model , or also as explained in Johannes Eber, "Anatomische Brillenanpassung", Verlag Optische Fachveroffentlichung GmbH, 1987, page 23 ff.
[0123] Figure 5 shows a detailed flow chart of an implementation of the method in Figure 4.
[0124] In steps 50 to 53 of Figure 5, the input data for the method are made available. In step 51 a frame manufacturer creates a parametric frame model for an eyeglass frame. The parametric frame model of step 51 can, in this case, be transferred to a standard unitary format, which is used in the method according to the invention, when the data is provided by the eyeglass manufacturer in a CAD (Computer Aided Design) format. ) owner.
[0125] In addition, a data reduction (for example, a reduction in the number of triangles or voxels in the 3D model) or a data compression can occur with the adjustment of conventional compression methods.
[0126] In step 50, the frame manufacturer creates specific adjustment requirements for that parametric frame model, which can take into account, as explained, aesthetic aspects in the frame adjustment.
[0127] In steps 52 and 53, a 3D model of the person's head is created and analyzed. In step 52, in this case, the model with a 3D measurement system is created first, particularly with the camera mechanism shown in Figure 2. Other measurement systems, such as 3D head scanners, can also be used. Examples of such head scanners can be found at http://cyberware.com/products/scanners/ps.html or http: //www.3d- shape.com/produkte/face_d.php, respectively, on June 8, 2017. In step 53 points or zones are then identified as features in this head model, for example, points or features, as they are also used in the state of the art explained at the beginning.
[0128] In step 54, then the frame is adjusted according to specific requirements, corresponding to step 40 of Figure 4. As an initial value for adjustment, in step 54, a nominal position and frame orientation can also be established with glasses. As a nominal position and nominal orientation, which can serve as an initial value for adjustment, it can serve a position by means of metadata, as in the patent application EU 17 173 929.5 with predefined standard parameters for the parametric frame model. Alternatively, the nominal position can be calculated, in some cases, from specific adjustment requirements. The specific adjustment requirements define, for example, the preferred location of the rim of the frame relative to the centers of the pupils in the xz plane; the distance to the apex of the nominal cornea (for example, 12 mm) defines the location in the direction of the y axis. The forward tilt as part of the frame orientation in space, that is, the angle around the x-axis, can be fixed at a nominal value of, for example, 9 degrees. This can also be an integral part of the specific adjustment requirements.
[0129] In step 55, the frame is then adjusted to the anatomical circumstances of the head. In that case, parameters that have not yet been adjusted in step 54, that is, parameters that are still free, are still adjusted.
[0130] In step 56 there is a virtual placement and rendering, and in step 57 there is a manual adjustment. Virtual placement and manual adjustment occur, in this case, as already described with reference to Figure 3, numerical references 33 to 310.
[0131] In step 58 there is a transmission to an order system of the manufacturer of frames, corresponding to step 311 of Figure 3.
[0132] The use of specific adjustment requirements for the frame and the corresponding adjustment will now be explained in more detail with reference to Figures 6 to 10.
[0133] Figure 6 shows different features of the face, which fit as features and points on the face for specific fit requirements of this type. In other words, in the adjustment requirements, in an example of such a modality, a target position or a target range of characteristics of the spectacle frame in relation to points of the face of that type is indicated. Such facial features are also explained in Johannes Eber, "Anatomische Brillenanpassung", Verlag Optische Fachveroffentlichung GmbH, 1987, page 17ss.
[0134] Examples are:
[0135] 1. Eye position, particularly the centers of the pupils (point of intersection of line L2 with lines LB in Figure 6). The L2 line also designates the pupil axis.
[0136] 2. Measure of the eye box, that is, measure of a rectangle, which is outlined around the eyes - position of each rectangle, width and height of the rectangles.
[0137] 3. Position of the nose corresponding to lines LA and L3 in Figure 6.
[0138] 4. Face width and temple position corresponding to the LD lines in Figure 6.
[0139] 5. Face height between lines L1 and L5 in Figure 6, as well as the chin line (line L5) in Figure 6.
[0140] 6. The radius of curvature of the chin area, that is, the part of the chin that touches the L5 line.
[0141] 7. Position of the eyebrows, with the L1 line in Figure 6 representing a central axis of the eyebrows and the LC lines representing an external delimitation of the eyebrows.
[0142] 8. Position of the mouth correspondingly to line L4 of Figure 6.
[0143] The above characteristics can be identified through an approach, as described later, through a parametric head model or also through image analysis methods (image recognition) and / or through mechanical learning in images captured by the camera mechanism of Figure 2 and, thus, its position can be determined in the 3D model of the head. A possibility for the automatic recognition of such characteristics is also described in V. Kazemi, J. Sullivan, "One millisecond face alignment with an ensemble of regression trees." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2014.
[0144] In the following description, designations such as left eye, right eye, left half of the face or right half of the face must be understood from the point of view of the person to whom the glasses are fitted.
[0145] Figure 7 shows a detailed method for adjusting the eyeglass frame based on the fit requirements, that is, a detailed example for step 40 in Figure 4 or step 54 in Figure 5, together with availability of the data.
[0146] In step 70 in Figure 7, adjustment requirements for a parametric frame model are made available, which are read in a computing mechanism in step 73, in order to be able to use them in the represented method. The tuning requirements are stored, for example, as text files, as an xml file or JSON file.
[0147] In step 71 a parametric frame model is made available, to which the adjustment requirements of step 70 are assigned. The parametric frame model can be assigned metadata that designates, for example, certain zones or points of the frame model. Such metadata for a frame model is also described in EU patent application 17 173 929.5. This parametric frame model is read in step 74. In step 77, the parameters of the parametric frame model resulting from the reading in step 74 and their value ranges are made available for subsequent optimization. Finally, in step 72 a 3D model of a person's head, which must be fitted to the glasses frame, is made available with associated metadata, which is read in step 75.
[0148] In step 76 there is a syntactic analysis of the adjustment requirements. Syntactic analysis means a decomposition and conversion of input data into a format suitable for further processing. A parser is a corresponding mechanism (usually implemented by a computer program), which performs such a parse. More detailed details can be found in the Wikipedia article "Parser", of 19 May 2017.
[0149] The adjustment requirements are translated particularly into a format that is suitable for a subsequent optimization method. The adjustment requirements may in that case, as explained, target sizes and / or accessible zones, particularly for distances between spectacle frame characteristics and head characteristics, for example, a distance between the upper frame edge and the eyebrows, a distance between the edge of the frame of the upper edge of the eyes, a distance from the bottom edge of a lower edge of the eyes or a relative location of the pupils of the edge of the frame. In addition, calculated derived feature distances, i.e. points or zones, which are derived from various head and / or frame features, can also be used. Derived characteristics of this type are also called auxiliary characteristics.
[0150] An example of such an auxiliary feature is shown in Figure 8. Figure 8 shows a head 80 with an eyeglass frame 81. With 80 an imagined circle with a radius of the semi-width of the face and midpoint on the edge is designated bottom of the nose. yUN designates, in Figure 8, the lower edge of the nose, yUK designates a lower edge of the chin, and yUOD designates a lower edge of the eyes. yG designates the width of the face, that is, the distance of the D lines of Figure 6. An example of an auxiliary characteristic derived yH, which is defined with the help of a term in the adjustment requirements 70, which is recorded during the analysis syntactic in step 76, is yH = (yUK - (yUN - 0.5 • xG)) / 0.5 • xG
[0151] This yH value represents a deviation of an arithmetic lower edge of the chin of an ideal face in relation to a real lower edge of the chin as a relation to the half-width of the face and is a measure for the vertical length of the face below the nose. Such an auxiliary feature can be used to establish the proportions of the lower frame edges of the spectacle frame. In this way, it can be taken into account that the length of the face in a vertical direction can also have an influence on the aesthetic impression that the spectacle frame produces and, thus, the specific adjustment requirements may predefine a ratio of the size and / or shape of the eyeglass frame for the yH parameter.
[0152] Another example of an adjustment requirement is a position of the pupils within a descriptive frame box. This is shown in Figure 10D. Figure 10D shows the spectacle frame 81 with a descriptive case of the frame 102, here for the right eye.
[0153] The height of the pupil's vision (height of the pupil on the lower edge of the frame) is designated with y, the horizontal location of the pupil is designated with x. The width of the box 102 makes Δa and the height of the box makes Δb. The adjustment requirement can then provide, for example, that the pupil may be located, in a horizontal direction, between the center of the box and the nasal golden section, that is, Δa • 3.82 <x <Δa • 0.5. The golden section means, in this case, that the relationship between x and Δa x is equal to the relationship between Δa x and Δa, which is the case for x = Δa • 3.82. Eye positions that are closer to the inside of the frame edge than this golden section are generally interpreted as less aesthetic.
[0154] A similar requirement may establish the position of the eyes in the vertical direction, namely that the pupil is located in the vertical direction between the center of the box of the box 102 and the value for the golden section above the center, that is, Δb • 0 , 5 <y <Δb • 0.618.
[0155] The adjustment requirements can also be made available directly as a calculation formula, and the variables in the calculation formula are then the characteristics described above. In other words, the frame parameters can be indicated directly in the specific adjustment requirement as a term, or they can be determined iteratively through an optimization polish. In the latter case, an adjustment quality is optimized, which is defined with the help of the terms; the terms set goals - but those goals were, in the general case, not achieved; therefore, for example, an expression of the form "target size = term" would contribute, in the sense of an optimization, for example, in the sense of the methods of the smaller squares only for the quality of the fit, but it would not be directly fulfilled.
[0156] The parsing in step 76 occurs, in particular, for the auxiliary characteristics addressed, for target sizes and calculation rules for them and optionally for a quality value such as scalar size, which is present, for example, in the form of a weighted quadratic sum of the deviation from the target sizes and may eventually present an additional punitive term, as described above.
[0157] In step 79, a list of syntactic trees for the terms of step 76 is then created.
[0158] Correspondingly, in step 78, for the head model, position, orientation and dimension are determined for values, such as position of the center of the pupil, position and dimension of the eye (for example, from a descriptive right corner of the eye ), position, orientation and size of the nose, position, orientation and position of the eyebrows and / or position of the chin.
[0159] In step 710, the tree terms for auxiliary characteristics are evaluated, that is, they determine which auxiliary characteristics are present, and in step 711, values for these auxiliary characteristics are determined, for example, for the yH value explained above . In step 712, an optimization step then occurs. Here, frame parameters of the parametric frame model are varied and the terms are evaluated, until target sizes have been reached in step 713. Hence, in 714, a set of parameters is given for a part of the frame parameters, which were adjusted based on the fit requirements. These parameters are particularly parameters that have an aesthetic effect, for example, scaling of the spectacle frame, tilting forward of the spectacle frame and / or a shape of the frame edge in the case of a variable frame edge. Other parameters, for example, platelet angle or a length of spectacle rods or a bridge width are left at default values, which are pre-defined by the manufacturer. These are then adjusted during anatomical adjustment (for example, step 41 in Figure 4).
[0160] The optimization polishing may also comprise a virtual placement, for example, as described in EU patent application 17 173 929.5. Through the previous steps, including the adjustment of the parameters of the parametric frame model, a convergence of the optimization for an optimum glasses fit is ensured. During the virtual placement, the result is, on the one hand, the parameters of the mathematical displacement (6 degrees of freedom, see the Wikipedia article, "Bewegung (Mathematik)", of May 22, 2017), representable, for example, as a rotation matrix and translation vector and, on the other hand, the frame flexion parameters. The latter is generally a unique parameter for the angle the ear support point travels during flexion. This corresponds to the virtual placement, as described in EU patent application 17 173 929.5. The result of the virtual placement described there are the rotation and translation of the frame and the parameters of the deformation of the rods.
[0161] After placement, all the specific characteristics of the frame are available in the head coordinate system. For this, the mathematical movement is applied to the characteristics. For example, the position and orientation of the right and left platelets of the individualized frame are calculated - that is, of the frame corresponding to the parametric frame model with adjusted parameters. Ideally, this position and orientation should coincide with the position previously calculated in the adjustment step. specific parameters for the frame, in which the corresponding characteristic of the nose flap has been harmonized with the characteristic on the frame, as will be explained more concretely later. However, it may happen that the method of virtual placement, due to limitations of individualization in the nose area, does not result in the same result when determining the position of the frame as the adjustment method. This can be conditioned, for example, by asymmetries of the real nose, in connection with a symmetrical nose support of the frame. Generally, the positions should, however, only be very slightly different. In the case of slight differences (for example, distance from the midpoint of the platelet less than 1 mm), this can be ignored. In the case of major differences, the new position after the virtual placement may trigger a new method of adjusting the parameters to be determined based on the specific adjustment requirements of the frame. A return is also possible in the form of an indication to the user of a possible compatibility of the frame model.
[0162] Figures 10A to 10C illustrate this positioning of the eye inside box 102 for different distances from pupils PD1 (Figure 10A), PD 2 (Figure 10B) and PD 3 (Figure 10C) in head 80, with PD 1 being a relatively small pupil distance, PD 2 is a medium pupil distance and PD3 is a relatively large pupil distance. For aesthetic adjustment, in the case of Figure 10A, an outer edge of the shape of the frame 100 is thickened and formed with dominant jaws, to preserve, for example, the condition of the golden section. The jaws are the outer part of the middle part of the spectacle frame, the inner part of which is called a bridge. The changed parameter is, therefore, the frame format. In the case of Figure 10C, a dominant zone or a dominant bridge is selected, possibly in connection with a wider bridge width, to obtain an unwanted aesthetic impression.
[0163] Figure 9 shows examples of adjusting parameters based on adjustment requirements to obtain a desired aesthetic effect. Figures 9A to 9C show, in this case, a staggering effect of the glasses frame 81. In Figure 9A, a very small frame is placed on the person, which is too small from an aesthetic and modern point of view. In Figure 9C the frame is very large. In Figure 9B the frame is medium in size. To ensure an aesthetically suitable glasses size, the adjustment requirements may, in this case, stipulate distances from the frame edge of the face edge and / or the eyebrows.
[0164] Figures 9D to 9F show an influence of the width of the bridge. The tip width is configured, in the modality example described here, during anatomical adjustment, to ensure an anatomically correct fitting of the glasses frame on the nose, which is explained in more detail below. However, it can also alter the aesthetic impression, which can still be taken into account during anatomical adjustment. In Figure 9D, a small bridge width b1 is selected. Here the frame rests too high due to a collision with the back of the nose. In Figure 9E the width of the bridge has been slightly extended to a width of the bridge b2. In this way, the glasses frame sits a little lower and more harmoniously. In the case of Figure 9F, the bridge width has been further reduced to a b3 value. Here, during anatomical adjustment, it can be guaranteed that the pupils are within a predefined range relative to the edges of the frame, for example, based on the golden section.
[0165] Thus, with the help of adjustment requirements and by splitting into an adjustment based on the adjustment requirements, followed by an adjustment to the anatomy of the head, it can be ensured that an eyeglass manufacturer's specifications, which are particularly aesthetic character, can be fulfilled.
[0166] In the case of the method referred to above, and also in other methods for adjusting glasses, for example, in the method described in patent application EU 17 173 929.5 or in some of the methods, which were explained as prior art , the position of certain points on the 3D model of the head is required and / or metadata is required, which characterize certain zones for adjusting glasses, such as a support point or an ear support zone. One possibility is to determine points or zones of this type manually or using sample recognition methods. Another possibility is now explained with reference to Figures 11 to 15.
[0167] Figure 11 shows a method for establishing measurement points on the 3D model of the person's head, according to an example of modality. Measurement points are points that can be used for the method described above, such as points that describe features of the face, such as ears, eyes, eyebrows and the like.
[0168] In step 110, a parametric head model with measuring points is available. A parametric head model is, in this case, a parametric model, which describes a head. By changing the parameters of the parametric model, the shape of the head described by the head model changes. The term parametric head model, as used herein, also includes models that describe only a part of the head, for example, only the parts that are necessary for an eyeglass fit (particularly the eyes, nose and ears). An example of a parametric head model will be explained later, with reference to Figures 13A and 13C. On this parametric head model, measurement points are established, for example, through manual selection. Examples of such measurement points will also be explained later, with reference to Figures 13A and 13C.
[0169] In step 111 the parametric head model is then adjusted to the 3D model of the person's head. For this, conventional random optimization methods can be used, which adjust the parameters of the parametric head model, so that a deviation as low as possible between the parametric head model and the 3D model of the person's head is present (for example , using the smaller squares method or the method in the article by J. Booth et al. cited above). In step 112 the measurement points are then transferred based on the adjustment to the 3D model of the person's head. In other words, the position of the measuring points on the adjusted parametric head model is used to establish corresponding measuring points on the 3D model of the head. This can occur through a projection of the parametric head model onto the 3D model of the head, for example, by using a section point of a normal vector, that is, of a vector perpendicular to the measurement point in the parametric head model , with the 3D model of the head. In exact models, the position of the measuring point can also be used directly in the parametric head model as a position in the 3D model of the head.
[0170] In this way, measurement points for essentially random 3D models of random heads can be determined, with the measurement points having to be established only once in the parametric head model.
[0171] Figure 12 shows a detailed method, which takes advantage of a parametric head model to establish measurement points in a 3D model of a person's head, incorporated in a method for the virtual adjustment of glasses. Instead of the method for virtual adjustment of glasses in Figure 12, the methods explained previously with reference to Figures 1 to 10 can also serve as a possibility of application for the method of Figure 11.
[0172] In Figure 12, a parametric frame model with free parameters is available in step 120. In the modality example in Figure 12, the free parameters are used for anatomical adjustment. In other examples of modality, an adjustment can also occur, as explained above, by means of specific adjustment requirements of the frame.
[0173] In step 121, a parametric head model is available. The parametric head model can be a face model or head model determined thanks to a principal component analysis (PCA), as described, for example, in A. Brunton, A. Salazar, T. Bolkart, S. Wuhrer, "Review of Statistical Shape Spaces for 3D Data with Comparative Analysis for Human Faces", Computer Vision and Image Understanding, 128: 1-17, 2014, or also a head model, as described in J. Booth, A. Roussos, S . Zafeiriou, A. Ponniah und D. Dunaway „A 3D Morphable Model learnt from 10,000 faces", 2016 IEEE Conference on Computer Vision and Patent Recognition (CVPR), Las Vegas, NV 2016 pages 5543 to 5552 issue date: 10.1109 / CVPR In step 122, a 3D model of the person's head is made available, which can be created, for example, with the camera mechanism of Figure 2.
[0174] In step 123, measurement points are determined on the parametric head model. An example of such a 3D model of at least part of the face is shown together with coordinating axes in Figure 14.
[0175] In step 123, measurement points are determined on the parametric head model. For this, a so-called standard head of the parametric head model is provided. A standard head is a head in which the parameters of the parametric head model take on predefined default values. In the case of a head model based on a principal component analysis, this can be, for example, an average head, which corresponds to a first component of the principal component analysis.
[0176] In step 123, measurement points are established on the parametric head model. This can happen manually by establishing points. An example of such an establishment is shown in Figure 13A. Here, in a standard head 130 of the parametric head model, a plurality of points is established, for example, angle of the mouth, tip of the nose, points on a forehead line, points of eyes, base of the nose and points on flaps of the nose. Another example is shown in Figure 13C. A triangle 132 is marked here, that is, three points on a nose flap of the head model 130.
[0177] In step 124 a parametric head model is fitted to the 3D model of the person's head using an adaptation method. An adaptation method is a method in which parameters of the parametric head model are determined, so that the parametric head model is adjusted as accurately as possible to the 3D model of the person's head, for example, according to the criteria of the smaller squares. Steps 123 and 124 can take place in a random sequence. Step 123 needs to be performed only once before the method is performed, so that the measurement points determined in each method implementation can be used for different 3D models of heads of different people and different parametric frame models.
[0178] In step 125 the measurement points are then transferred to the adapted parametric head model. In other words, the position of the measuring points is determined in the adapted head model. For this, the same transformation that is used to arrive, from the standard head model in which measurement points were determined in step 123, to the adapted parametric head model, is applied essentially to the measurement points, for example, as described in the article by J. Booth et al. mentioned above. In step 126, the measurement points are optionally transferred to the 3D model of the head. Whether step 126 is used or not depends on the accuracy of the model used, that is, how exactly the adapted parametric head model corresponds to the 3D model of the person's head. If, for example, the mean square deviation is less than a threshold value, step 126 can be omitted. The transfer of the measuring points from the parametric head model adapted to the 3D model of the person's head can occur through a projection, in which a normal vector is determined by a respective measurement point in the adapted head model, and the The section of this normal vector with the 3D model of the person's head is then used as the corresponding measurement point in the 3D model of the person's head. Examples are shown in Figures 13B and 13D. In Figure 13B the points in Figure 13A are projected on a 3D model 131 of the person's head, and in Figure 13D the triangle 132 in Figure 13C is projected as triangle 132 ‘in 3D model 131.
[0179] This projection works reliably on many face models, as parametric models often have a high smoothness, particularly a higher smoothness than a typical 3D model of the head, as shown in Figure 14. The surface smoothness in this case, it can be defined as a measure of local deviation from normal vectors. Alternatively, as a measure, the local deviation of the point cloud of the 3D model of the head can also be defined by an approximate polynominal surface, for example, respectively in local zones of 5 mm in diameter. Polynomial surfaces are often infinitely differentiable and are therefore called "smooth" in differential geometry. A local smoothing by means of "moving least squares" (MLS), which can be applied in examples of modality, is described in http://pointclouds.org/documentation/tutorials/resampling.php, of 8 June 2017.
[0180] In addition, (not shown in Figure 12) a manual step can be used to mark other measurement points on the 3D model of the head. These can be, in particular, points that are not easily registered with the 3D model, for example, parts of the person covered by hair. This may be the case, in particular, for the ears. Therefore, these points are not exactly identifiable in the 3D model of the person's head and can be added manually. An example of such a measuring point is a point of support for the spectacle rods at the base of the ears.
[0181] In step 127 characteristics are calculated based on measurement points (in measurement points in the adapted head model, if step 126 does not apply or in the transferred measurement points, if step 126 is performed). These features, also called measurement features, are based on groups of measurement points and define, for example, a head area.
[0182] The characteristics can be determined by means of direct calculation (for example, 3 non-collinear points in space clearly define a plane, whose normal vector can be calculated using a Cartesian product from the normalized differential vectors; 4 non-coplanar points define a sphere, 5 non-coplanar points define a cylinder) or by approximating a geometric primitive (points, lines or surfaces), such as a plane or ball or a cylinder, to certain measurement points. The characteristic is then determined through the parameters of the geometric primitive, in the case of a plane, for example, through a normal vector and receiving point of the plane, in the case of a ball through the midpoint and radius of the ball, etc. The following are examples of such characteristics, which are calculated in step 127: - left or right nose flap
[0183] For the left or right flap of the nose, a plane (for example, corresponding to triangle 132 'of Figure 13D) defined by approaching a small area of the model, in the area of the nose support or of the platelet areas (for example, with a diameter of 6 mm). From the plane's location the horizontal and vertical nose flap angles are given. In this case, the plane at the midpoint of the nose support area is cut with the coordinate axes and, respectively, the resulting angle is measured. If, for example, three points are marked, corresponding to triangle 132, on each nose flap in Figure 13C, the plane can be calculated from the three points. In the case of more than three points, the plane can be calculated using an adjustment method, for example, by decomposing main components into the number of points, or through an adjustment with the help of the smaller squares method. An individual plane is represented, as mentioned above, by a point (x, y and z) on the plane and by a normal vector (nx, ny, nz) by these points, where x, y and z are Cartesian coordinates. Both nose flaps together can therefore be represented as double-fold, that is, as 12 values (2 points and 2 normal vectors), for example, as (x [N, OD], y [N, OD], z [N, OD], nx [N, OD], ny [N, OD], nz [N, OD], x [N, OS], y [N, OS], z [P, OS], nx [N, OS], ny [N, OS], nz [N, OS])
[0184] In this case, the N index designates the nose, the OD index designates the right eye (oculus dexter) and the OS index designates the left eye (oculus sinister). - forehead curvature
[0185] Here, a section of a circular curve can be adjusted in space to measurement points on the forehead, as shown in Figures 13A and 13C. Parameters of this adjustment are midpoint, radius and normal vector of a plane, which is located in the circle. This adjustment can take place in two steps. First, an adjustment of a plane is carried out, as described above for the nose flaps, and a circle is then still adjusted in the plane. This adjustment of the circle can happen, for example, with the method of the smaller squares or in any other conventional adjustment method. - eyebrow and / or cheekbones
[0186] Here, an S spline surface is fitted in a field around the eyebrows and / or in a field around the cheekbones (see the Wikipedia article "Spline", 23 May 2017) or a bivariate polygon (see, for example, https://en.wikipedia.Org/wiki/Polynomial#Definition ^ "polynomial bivariate", of 8 June 2017), the measurement points in the eyebrow field and in the apple field the face. In a spline representation S (c1, .., cn): (x, z) ^ y coefficients (c1 ..., cn) of the spline function S are determined, so that for a number of measuring points { (x1, y1, z1),., (xm, ym, zm)} in the corresponding field (eyebrows or cheekbones), an error F is minimal in the mean square root, that is, error F has the following configuration F (c1, ..., cn) = ∑i = i ... m (yi-S (c1, .., cn) (xi, zi)) 2
[0187] In this representation, it is assumed that, subsequently, the method of placing the frame occurs through displacement parallel to an xy axis with a y value respectively fixed in the coordinate system of Figure 14. If it is necessary to achieve a minimum distance from an edge posterior to the frame of the 3D model of the head using the adjustment method, this distance value can be previously defined as offset on the spline surface. A touch by matching the y-values can then be detected (since the y-value is previously stored as an offset). For this, during the posterior adjustment of the glasses frame, each vertex of the posterior edge of the frame can be inspected, and a respective vertex granted by the coordinates (x, y, z) is inspected with respect to the difference Δy = y - S (c1 , ... cn) - (x, z) examined. During the detection of a touch or immersion of the vertex in the model, a position of the spectacle frame can then be adjusted or the edge of the spectacle frame can be modified. - base point of the ears, which serves as a support point for the spectacle rods
[0188] For this purpose, an individual point on the head model can be used, that is, no measuring points need to be grouped here. In other embodiments, an ear support curve can be determined, as described in EU patent application 17 173 929.5. If a model without ears modeling is used (see above), for example, a pure face model, or the ears were covered during the creation of the 3D model of the person's head, that base point of ears can be generated from another way, for example, by mechanical learning in images, which have been used for the creation of the 3D model of the head, and for that, a trained feature detector can be used to detect the base point of the ears in the images. These points detected in the 2D image are projected, in another step, on a 3D model of the head. Information about such projections can be found in the basic literature on projective geometry and camera calibration, for example, Hartley and Zisserman, "Multiple View Geometry in Computer Vision", 2000, starting on page 7 for image pixel representation like straight lines in space; projection on a 3D model in space as calculation of the frontmost section point in the triangulation network with straight lines, also called "Ray-Casting", see also, for example, the software library "vtk", function "vtkModifiedBSPTree :: IntersectWithLine ". Alternatively, such a point can also be determined manually, as described above.
[0189] In some examples of modality, certain points, such as eye position or pupil position, can also be determined with a separate method, for example, images captured by pupil detection and corneal detection with the Figure camera 2. Determinations of these types are described in patent applications EU 17 153 558.3 and EU 17 153 559.4.
[0190] Based on the characteristics thus calculated in step 127, in step 128 the frame parameters of the parametric frame model are then calculated. The following is an example of this calculation. However, the features can also be used for the frame adjustment described above, based on specific adjustment requirements or for virtual placement, as described in EU 17 173 929.5 patent application.
[0191] Generally, for adjustment, the characteristics are evaluated in combination, in relation to the relative location and / or other properties, such as angle or curvature. The following are some examples for calculating the frame parameters in step 128. These can also serve as an example for the anatomical adjustment of step 41 in Figure 4. - bridge width
[0192] The width of the bridge is defined in DIN EN ISO 8624: 2015-12, annex A, and occurs from the relative position of the platelets, since, in the case of a wider bridge width, the platelets are more widely spaced and, in the case of a narrower bridge width, the platelets are closer together. In the case of a spectacle frame without platelets, generalized platelets are defined as special zones, which are provided as zones of contact with the nose. The width of the tip is given as a distance from the midpoints of these generalized platelets. Thus, the width of the bridge can correspond to a distance of midpoints of triangles on both flaps of the nose, correspondingly to the triangle 132 ‘of Figure 13D. In this case, the geometric focus can be assumed as the midpoint of the triangle, that is, the midpoint of the bisectors.
[0193] Figure 16 shows, for illustration, a perspective view of a parametric frame model with platelets 160 (in that sense) and the width of the bridge 161. - relative position and angle of the platelets
[0194] This adjustment is explained in Figure 15. Here are shown nose flaps as a cross section. This is represented by a curve 150, and a plate 151 is fitted.
[0195] Each of the two nose flaps can be adjusted using a plane, based on the respective plate (tangential plane). This platelet plane, as described above for other planes, is approximated through a receiving point (xP, yP, zP) and a normal vector (nx, ny, nz). The receiving point can be assigned, in particular, as the platelet midpoint. In platelets in the classic sense, that is, in metal frames, this midpoint is defined, for example, through a projection of the focus of the tag on the outer side, that is, the contact surface of the tag with the nose - the midpoint of the nameplate can also, as a pre-defined point, be part of the parameterized frame model - that is, this point is provided together with the model. In the case of plastic frames without displaced plates, the part of the frame, which is designed as a contact surface for the nose (160 in Figure 16), called the nose support or here, in general, by the plate. In this way, the two platelets can also be represented as double-fold, and the representation, in the present example of modality, occurs in the local coordinate system: (x [P, OD], y [P, OD], z [P, OD], nx [P, OD], ny [P, OD], nz [P, OD], x [P, OS], y [P, OS], z [P, OS], nx [P, OS ], ny [P, OS], nz [P, OS]) where the P index represents the nameplate.
[0196] The location of the platelets implies, as explained above, also the width of the bridge.
[0197] In this representation of the platelets, the origin of the coordinates and the orientation of the coordinate system can be freely selected, since the double-fold is capable of being transferred through a common translation illustration at the receiving point and an illustration of common rotation and normal vectors for any coordinate system. The prerequisite is that in the parametric frame model, all parameters can be freely selectable from the above two-fold. In practice, in a parametric frame model, the parameters are limited and there are maximum and minimum values for the individual parameters of the parametric frame model (for example, a frame cannot be manufactured with a random size or with a bridge width at random larger or randomly smaller). In any case, both the platelets and, as discussed above, the flaps of the nose, can be represented as double-fold.
[0198] Normal vectors can be represented, respectively, by two angles theta and fi in space, instead of Cartesian coordinates as above (essentially a representation in polar coordinates, as the length (radius) of normal vector 1 is selected : (nx, ny, nz) = (sin (fi) * sin (theta), cos (fi) * sin (theta), cos (theta)).
[0199] In this way, a total of 10 degrees of freedom are given for the platelets and thus also for the bridge of the nose) together, with a representation being obtained as a tenfold: (x [P, OD], y [P, OD], z [P, OD], thetaOD, fiOD, x [P, OS], y [P, OS], z [P, OS], thetaOS, fiOS) = z [P, OS]
[0200] The interdependence between the width of the nose bridge and the position of the platelets is visible from Figure 15. If the nose bridge is enlarged, the distance between the receiving points of the left and right platelet planes increases accordingly, and vice versa.
[0201] A reduction in the number of parameters occurs when it is assumed that the bridge and the platelets are symmetrical to each other. With the yz plane in Figure 14, the following symmetry plane applies: i. X [P, OS] = - X [P, OD] ii. y [P, OD] = y [P, OS] and z [P, OD] = z [P, OS] iii. theta [P, OD] = theta [P, OS] and fi [P, OD] = - fi [P, OS]
[0202] Then they are given as free parameters (w, yP, zP, theta, fi) with theta = theta [, oD] = theta [, os] and fi = fi [, oD] = - fi [ , you]. w is, in this case, the width of the bridge, with x [, oD] = w / 2 and x [, oS] = -w / 2 being applied. Thus, in the symmetrical case, 5 free parameters are present, with which the parametric frame model can be adjusted. Depending on the frame, less degrees of freedom may be present, or degrees of freedom may be limited by means of adjustment requirements, as explained above.
[0203] To adjust the parametric frame model to the 3D model of the head, the planes of the platelets can be selected so that they coincide with the planes of the nose flaps, that is, in the general case, the duodecuples for the platelets match with the double for the nose flaps.
[0204] As a limitation, for example, the position of the bridge or platelets can be fixed in the local coordinate system of the frame (ie, the values yP, zP are fixed), or for theta and fi a fixed reference can be selected , for example, linear, in relation to each other, so that theta and fi are not selectable independently of each other.
[0205] In the case of a nose flap reduced from frame parameters, for example, in the symmetrical case mentioned above, a medianization can be used. If, for example, the theta [P, OD] and theta [P, OS] angles corresponding to the nose flaps are different, an average value can be used. If the angles are more pronouncedly different from one another than a limit value, an alert can be issued, so that the symmetrical shape of the frame results in unfavorable wear properties here. To assess how unfavorable these wear properties are, a quality measure that designates the quality of anatomical fit can be used. Such a measure of quality can be calculated on the basis of the distances referred to above the glasses frame of the head areas, whereas different distances with a different weighting can be incorporated into the measure of quality.
[0206] Depending on the parametric frame, the number of free parameters can be further reduced, for example, to two parameters in the area of the nose support, namely the width of the bridge and a parameter for the web angle. The soul angle is explained, for example, in Johannes Eber, "Anatomische Brillenanpassung", Verlag Optische Fachveroffentlichung GmbH, 1987, page 26, image 24 of the soul angle. - pantoscopic frame angle
[0207] Also, by means of the characteristics, the pantoscopic angle of the frame (also called forward tilt angle) can be calculated or adjusted. In examples of modality in which specific adjustment requirements for the frame are used, as explained above, the pantoscopic angle can already be configured during this adjustment (step 40 in Figure 4). This can then be further adjusted in step 128 of Figure 12. To do this, a distance from the edge of the frame is calculated (for example, the rear edge of the bottom edge of the edge of the frame, bottom left or right corner in a front view of the frame ) of the cheek surfaces mentioned above, which may be represented by a spline surface. The pantoscopic angle is then changed, so that a minimum pre-defined distance, for example, 2 mm, is ensured. - rod length
[0208] The length of the rods is calculated in step 128, when the seat of the frame has been established on the nose, for example, on the platelets mentioned above. To configure the length of the stems (as long as this is a free parameter of the parametric frame model), a forward support point of the stems is covered by the points of the ear base mentioned above.
[0209] In step 129, frame parameters calculated in step 128 are then applied to the parametric frame model. In step 1210 there is then a virtual placement and rendering, as described with reference to step 56 of Figure 5. In step 1211, yet another optimization can take place, for example, an optimization occurs, as described in the document US 2016/0327811 A1 referred to at the beginning, or a manual adjustment, as described in step 57 of Figure 5. In step 1212, a transfer to the ordering system takes place. Other parameters can also be selected, for example, a color of the middle part of the eyeglass frame, a color of the eyeglass rods of the eyeglass frame, material and color of the hinge of the eyeglass frame, engravings on the eyeglass rods of the eyeglass frame , design elements, applications on spectacle rods or the middle part of the spectacle frame. The requested eyeglass frame is then manufactured according to the parameters determined, for example, with an additive manufacturing method, as explained at the beginning.

权利要求:
Claims (9)
[0001]
1. Computer-implemented method for the virtual adjustment of glasses, which comprises defining first measurement points in a 3D model of a person's head (122), with measurement points being points in a model, which can be used to a subsequent adjustment of glasses, and fitting (128) a model of an eyeglass frame (120) to the 3D model of the head (122) based on the first measuring points, characterized by the fact that the definition of the first measuring points, characterized by the fact that defining the first measuring points comprises, fitting (124) a parametric head model to the 3D model of the person's head, and determining (125, 126) the first measuring points based on defined second measuring points in the parametric head model and in the adjustment of the parametric head model to the 3D model of the head.
[0002]
2. Method, according to claim 1, characterized by the fact that the second measuring points are defined in a standard head of the parametric head model, with the determination of the first measuring points comprising a transfer (125) of the second ones characteristics defined in the standard head for the adjusted parametric head model, corresponding to the adjustment.
[0003]
3. Method according to claim 2, characterized by the fact that the determination of the first measuring points comprises the use of the second transferred measuring points as the first measuring points.
[0004]
4. Method according to claim 2, characterized by the fact that the determination of the first measurement points comprises a projection (126) of the second measurement points transferred on the 3D model of the head.
[0005]
Method according to any one of claims 1 to 4, characterized by the fact that by combining (127) several first measuring points to form a feature, which defines an area of the 3D model of the head.
[0006]
6. Method according to claim 5, characterized by the fact that the combination comprises an adjustment of a geometric figure or a function to the various measurement points.
[0007]
Method according to either of claims 5 or 6, characterized in that the area of the 3D model comprises nasal flaps, a curvature of the forehead, eyebrows or a cheek area.
[0008]
Method according to any one of claims 1 to 7, characterized in that the spectacle frame model comprises a parametric frame model and that the adjustment comprises a determination of one or more parameters of the parametric frame model based on the first measuring points and / or the characteristic.
[0009]
Method according to any one of claims 1 to 8, characterized in that it still comprises a definition (123) of the second measurement points in a parametric head model (121).

类似技术:

---

公开号 | 公开日 | 专利标题

---

BR112020000106B1|2021-01-05|computer-implemented method for virtual adjustment of glasses

---

US11262597B2|2022-03-01|Method, device, and computer program for virtually adjusting a spectacle frame

---

BR122020012139A2|2020-10-27|computer-implemented process for adjusting glasses and device for data processing

---

KR20200006621A|2020-01-20|Methods, apparatus, and computer programs for determining near vision points

同族专利:

---

公开号 | 公开日

---

JP2021099504A|2021-07-01|

---

CN111033364A|2020-04-17|

---

EP3425446B1|2019-10-30|

---

EP3425446A1|2019-01-09|

---

EP3649505A1|2020-05-13|

---

JP6840889B2|2021-03-10|

---

CA3072117A1|2019-01-10|

---

KR102217887B1|2021-02-22|

---

EP3649505B1|2020-11-25|

---

WO2019007939A1|2019-01-10|

---

ES2845569T3|2021-07-27|

---

KR20200019757A|2020-02-24|

---

CN113985628A|2022-01-28|

---

JP2020525858A|2020-08-27|

---

BR112020000106A2|2020-07-14|

---

US20200233239A1|2020-07-23|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

JPS6118349B2|1975-12-22|1986-05-12|Hitachi Ltd|

---

WO2001088654A2|2000-05-18|2001-11-22|Visionix Ltd.|Spectacles fitting system and fitting methods|

---

FR2812506B1|2000-07-25|2002-12-20|Canon Kk|ALARM METHOD AND DEVICE DURING THE PROGRESSIVE DECODING OF A DIGITAL IMAGE CODED WITH A REGION OF INTEREST|

---

US6792401B1|2000-10-31|2004-09-14|Diamond Visionics Company|Internet-based modeling kiosk and method for fitting and selling prescription eyeglasses|

---

US7016824B2|2001-02-06|2006-03-21|Geometrix, Inc.|Interactive try-on platform for eyeglasses|

---

EP1495447A1|2002-03-26|2005-01-12|KIM, So-Woon|System and method for 3-dimension simulation of glasses|

---

DE10216824B4|2002-04-16|2006-03-02|Thomas Doro|Method and apparatus for constructing a custom goggle|

---

FR2971873B1|2011-02-22|2014-01-03|Fittingbox|METHOD FOR DETECTING A PREDEFINED SET OF FACE CHARACTERISTIC POINTS|

---

US20130088490A1|2011-04-04|2013-04-11|Aaron Rasmussen|Method for eyewear fitting, recommendation, and customization using collision detection|

---

US8733936B1|2012-01-30|2014-05-27|Ditto Technologies, Inc.|Fitting glasses frames to a user|

---

US9286715B2|2012-05-23|2016-03-15|Glasses.Com Inc.|Systems and methods for adjusting a virtual try-on|

---

FR2995411B1|2012-09-07|2014-09-19|Tipheret|METHOD AND DEVICE FOR PREPARING A GLASSES FRAME|

---

KR102207026B1|2013-08-22|2021-01-22|비스포크, 인코포레이티드|Method and system to create custom products|

---

FR3016051B1|2014-01-02|2017-06-16|Essilor Int |METHOD FOR DETERMINING AT LEAST ONE GEOMETRIC PARAMETER OF A PERSONALIZED FRAME OF EYEWEAR AND METHOD OF DETERMINING THE ASSOCIATED CUSTOM FRAME|

---

FR3016052B1|2014-01-02|2018-03-30|Essilor International|METHOD FOR DETERMINING A GEOMETRIC DEFINITION OF A CUSTOM OPTICAL EQUIPMENT|

---

FR3016050B1|2014-01-02|2017-12-08|Essilor Int |METHOD OF ADJUSTING A PREDETERMINED GLASS MOUNT FOR USE BY A DONOR|

---

US20150277155A1|2014-03-31|2015-10-01|New Eye London Ltd.|Customized eyewear|

---

US20150293382A1|2014-04-09|2015-10-15|Pro Fit Optix, Inc.|Method and System for Virtual Try-On and Measurement|

---

FR3021205B1|2014-05-20|2021-12-24|Essilor Int|METHOD FOR DETERMINING AT LEAST ONE BEHAVIORAL PARAMETER|

---

WO2016164859A1|2015-04-10|2016-10-13|Bespoke, Inc.|Systems and methods for creating eyewear with multi-focal lenses|US10685457B2|2018-11-15|2020-06-16|Vision Service Plan|Systems and methods for visualizing eyewear on a user|

---

EP3702831A1|2019-03-01|2020-09-02|Carl Zeiss Vision International GmbH|Dataset for use in a method for manufacturing an ophthalmic lens|

---

US11238611B2|2019-07-09|2022-02-01|Electric Avenue Software, Inc.|System and method for eyewear sizing|

---

WO2021062440A1|2019-09-24|2021-04-01|Bespoke, Inc. d/b/a Topology Eyewear|Systems and methods for adjusting stock eyewear frames using a 3d scan of facial features|

---

EP3809304A1|2019-10-18|2021-04-21|Carl Zeiss Vision International GmbH|Computer-implemented method for determining a parametric substitute model of a spectacle frame element, device and systems using such a method|

---

US11250572B2|2019-10-21|2022-02-15|Salesforce.Com, Inc.|Systems and methods of generating photorealistic garment transference in images|

---

EP3876026A1|2020-03-06|2021-09-08|Carl Zeiss Vision International GmbH|Method and devices for determining inclination angle|

---

JP2021149031A|2020-03-23|2021-09-27|ホヤ レンズ タイランド リミテッドHOYA Lens Thailand Ltd|Virtual-image generating device and virtual-image generating method|

---

EP3944004A1|2020-07-23|2022-01-26|Carl Zeiss Vision International GmbH|Computer-implemented method for generating data for producing at least one spectacle lens and method for manufacturing spectacles|

---

WO2022022765A1|2020-07-31|2022-02-03|Tribe Gmbh|Method and device for automatically determining production parameters for a pair of spectacles|

法律状态:
2020-10-20| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|

---

2021-01-05| B16A| Patent or certificate of addition of invention granted [chapter 16.1 patent gazette]|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 03/07/2018, OBSERVADAS AS CONDICOES LEGAIS. |

优先权:

---

申请号 | 申请日 | 专利标题

---

EP17179990.1|2017-07-06|

---

EP17179990.1A|EP3425446B1|2017-07-06|2017-07-06|Method, device and computer program for virtual adapting of a spectacle frame|

---

PCT/EP2018/067914|WO2019007939A1|2017-07-06|2018-07-03|Method, device and computer program for virtually adjusting a spectacle frame|

---