Computer-implemented method for personalizing eyeglasses frame elements by determining a parametric replacement model of the eyeglasses frame elements, and device and system using such method

Abstract

This invention relates to a computer-implemented method (10, 10', 10") for personalizing an eyeglass frame element (24) by adapting a parametric model of the eyeglass frame element (24) to the head of an eyeglass wearer. A parametric equivalent model of the parametric model of the eyeglass frame element (24) having at least one parameter is determined by: specifying a plurality of entities (30) of the parametric model in the form of implementing the parametric model using specific parameter values; determining at least one base entity (38) and at least one parametric deformation mapping of the at least one base entity (38) based on these specified entities (30); the at least one parameter... The deformation mapping maps the at least one base entity (38) to the entity (30) of the parameterized model, and the parameterized equivalent model is determined at least based on the at least one base entity (38) and the at least one parameterized mapping. Biometric data related to the head of the eyeglass wearer are determined, and at least one parameter value of at least one parameter of the parameterized equivalent model of the eyeglass frame element (24) is determined by optimizing a function that takes into account at least one surface point of the determined base entity (38) of the parameterized equivalent model of the eyeglass frame element (24) and the determined data (31) related to the head of the eyeglass wearer.

Description

Technical Field

[0001] This invention relates to a computer-implemented method for personalizing eyeglass frame elements by adapting a parametric model of the frame element to the head of an eyeglass wearer through determining a parametric equivalent model, the parametric equivalent model having at least one parameter of the parametric model of the eyeglass frame element. The invention also relates to an apparatus for personalizing and adapting a parametric model of an eyeglass frame element to the head of an eyeglass wearer, and an apparatus for representing and / or compressing a given entity of a parametric model of an eyeglass frame element. Furthermore, the invention relates to a computer program product having a computer program with program code for performing the method, and a system having an apparatus for generating personalized eyeglass frame elements or for grinding eyeglass lenses into personalized eyeglass frame elements. Background Technology

[0002] To date, centering measurement equipment for eyeglass frame elements has facilitated fully automated, computer-controlled centering measurements, the personalization of eyeglass frame elements as parametric models, and the adaptation of said frame elements to the head of the eyeglass wearer. For this purpose, 3D scanning methods are used to measure portions of the eyeglass wearer's head, and the resulting head model is stored in the random access memory or hard disk space of a computer unit. The extent and / or alignment of the eyeglass frame elements or their portions, as well as the distances and / or angles between their portions, are preferably varied in a way that the geometry of the eyeglass frame elements corresponds to that of the head model to adapt to the eyeglass wearer's head model.

[0003] Eyeglass frame components are typically stored in the memory of a computer unit as parametric models, such as CAD models, in a program-specific data format, such as STL, STEP, OBJ, or PLY files. Modeling programs, such as CAD programs like Creo, SolidWorks, Autodesk, FreeCAD, or OpenSCAD, can be used to generate such models.

[0004] However, these modeling programs typically do not include the ability to adapt parametric models of eyeglass frame elements to head models. Instead, they can only generate and store solid forms of parametric models of eyeglass frame elements. However, such solid forms are unsuitable for personalizing and adapting eyeglass frame elements because they do not contain any parameters and therefore cannot be modified or adapted to the head model of the eyeglass wearer.

[0005] Furthermore, parametric models generated based on modeling procedures are not suitable for use in systems independent of modeling procedures (e.g., fitting systems for eyeglass frame components) for the following reasons:

[0006] First, modeling programs typically do not offer the option to export and store the underlying parametric model of the eyeglass frame components. This is because the representation format of the parametric model used by the modeling program is usually designed only for internal representation and processing of data within the corresponding modeling program, rather than for systems independent of the modeling program. Second, the methods used to represent and use the parametric model within the modeling program are usually not publicly accessible, therefore the parametric model cannot be used without additional information.

[0007] Therefore, to personalize and adapt eyeglass frame components, it is necessary to make the parametric model available outside the modeling process. For this purpose, a so-called reverse engineering method can be used, which generates a parametric equivalent model for a given parametric model, independent of its modeling process. In this case, maximizing the automation of the time-consuming process of generating a modifiable parametric equivalent model for a given base model is particularly important.

[0008] A method for determining a parametric replacement model based on a parametric model is known from WO 2019 / 051243 A1. This method includes the following steps:

[0009] - Provides solid models of CAD components.

[0010] - Identify one or more geometric features of a CAD model, such as holes, flanges, pipes, walls, tension or load areas, and modify them based on rules and templates;

[0011] - Automatically create parametric equivalent models based on geometric features with geometric parameters (e.g., width, height, thickness, diameter, etc.);

[0012] - Calculate the modified entity of the CAD model based on the selection of geometric features and associated parameter values.

[0013] In this scenario, one or more geometric features, such as holes, flanges, pipes, and walls, are automatically identified and modified based on predefined rules and templates. This requires programming corresponding identification routines for each specific feature and defining geometric parameters for the modifications to each feature. Since the specified method is not specific to the eyeglass frame element, defining identification routines and parameters for each feature of the eyeglass frame element represents a significant overhead for the programmer. Furthermore, because the parametric equivalent model is not defined in a data-driven manner but based on programmer-defined rules and templates, this approach can easily lead to unrealistic parametric equivalent models. Moreover, only individual geometric features of the model are detected and parameterized, meaning only these are modifiable, not the entire object. Additionally, the computed parametric model has the same data format as the base model, so the use of the equivalent model depends on the modeling procedure.

[0014] For these reasons, this method is not suitable for, for example, personalizing and adapting complex models (such as eyeglass frame elements) with many different geometric features in adaptation systems.

[0015] The publication "CAD Model Creation from Dense Pointclouds: Explicit, Parametric, Free-Form CAD and Re-engineering" in the book *Advanced CAD Modeling* (Springer-Verlag, 2019, pp. 217-239) discloses a method for automatically reconstructing objects with free-form surfaces from point clouds in NURBS format. However, this process does not generate a parametric model of the underlying point cloud.

[0016] The publication "Automatic and Parametric Mesh Generation Approach," by Alan M. Shih, Sankarappan Gopalsamy, Yasushi Ito, Douglas Ross, Mark Dillavou, and Bharat Soni (2005), describes a method for generating optimized geometries for a given purpose by using parametric models based on parameter variations and simulations. However, no automated method is available for imported object geometries.

[0017] The publication “Development of Parametric Mesh Morphing Techniques”, Makoto Onodera, Ichiro Nishigaki, Yoshimitsu Hiro, Chikara Kongo, Transactions of the Japan Society of Mechanical Engineers Series C, 74, 2008, pp. 1894-1600, describes a method for identifying the geometric features of meshes in the form of planar, quadratic, or freeform surfaces.

[0018] US 2016 / 0327811 A1 describes a method for adapting eyeglass frames. However, the eyeglass frame here cannot be treated as a parametric model whose parameters are changed, but rather directly deformed. Only elastic deformation is conceived in this process. Nor can the amount of material be changed.

[0019] US 2016 / 336737 A1 describes the adaptation of eyeglass frames based on parameterizable frame models. However, this does not involve generating parameterizable eyeglass frame models from a given parameterizable frame model.

[0020] EP 2 746 838 A1 describes an adaptation system for virtual eyeglass frames based on a parametric model of the eyeglass frame. However, this parametric model is directly based on space curves and closed volumes and is not used to generate parametric equivalent models.

[0021] Therefore, the above methods do not allow for the automatic generation of a parametric equivalent model from a given parametric model. Summary of the Invention

[0022] Therefore, the object of the present invention is to facilitate the highly automated determination of parametric equivalent models of parametric models of eyeglass frame elements, wherein the parametric equivalent model has at least one parameter.

[0023] This objective is achieved through the present invention.

[0024] The first computer-implemented method according to the present invention for personalizing eyeglass frame elements by adapting a parametric model of the eyeglass frame element to the head of an eyeglass wearer includes the following method steps:

[0025] A parametric model is defined with multiple entities. At least one base entity and at least one parametric deformation mapping of the base entity are determined based on the specified entities, the at least one parametric deformation mapping mapping the at least one base entity onto the entities of the parametric model. In this process, a parametric equivalent model is determined at least based on the at least one base entity and the at least one parametric deformation mapping. Furthermore, biometric data related to the head of the eyeglass wearer are provided, and at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element is determined by optimizing a function that considers at least one surface point of the determined base entity of the parametric equivalent model of the eyeglass frame element and the determined biometric data related to the head of the eyeglass wearer.

[0026] In this context, biometric data related to the head of an eyeglass wearer is understood to refer to data describing the biological properties of the head, specifically dimensions such as head length, size, distance, and ratio, for example, interpupillary distance, bridge of the nose width, and / or interauricular distance, as well as surface points of the head, such as ear fulcrums, nasal fulcrums, pupils, and head models of the eyeglass wearer, such as 3D models, specifically head meshes, 3D reconstructions or point clouds, or models or dimensions of parts of the eyeglass wearer's head.

[0027] In the current context, a frame element describes a portion of the eyeglass frame, such as the temples, nose support area, bridge, connecting elements, or the front of the frame. However, a frame element can also refer to a frame element or a combination of segments of frame elements, such as a temple segment, or the entire eyeglass frame.

[0028] The present invention interprets the parametric model and parametric equivalent model of the eyeglass frame element as a three-dimensional representation of the eyeglass frame element in a computer unit, the representation including at least one parameter for adjusting the features and / or properties of the eyeglass frame element or a portion thereof, such as temple length or working angle of the nose support surface.

[0029] For example, a parametric model of an eyeglass frame component can serve as a CAD model. In the present context, a CAD model should be understood as a representation of a 3D object that can be processed by a computer unit, specifically one that can be read into and stored within the computer unit, such as as a file in the computer unit's hard disk space.

[0030] A parametric equivalent model is a parametric model used in place of another parametric model in processes such as the personalization and adaptation of eyeglass frame components, and thus replaces the other parametric model (base model).

[0031] In this context, parameters represent variable values ​​that can affect the characteristics and / or properties of eyeglass frame elements or parts thereof. Parameter values ​​represent specific numerical values ​​that can be used for these parameters.

[0032] This invention interprets an entity of a parameterized model or an entity of a parameterized equivalent model as a specific example of an implementation of the parameterized model or a parameterized equivalent model for selected parameter values. In this process, parameter values ​​are assigned to each parameter of the parameterized model or parameterized equivalent model.

[0033] This invention understands the basic entity as the selected or computed parametric model and the specific entity used to define the parametric deformation mapping.

[0034] Parametric deformation mappings are parameterized mappings, such as parameterized affine mappings, that act on the surface of a given base entity. By choosing specific parameter values ​​for the mapping's parameters, a specific mapping that alters the surface of the base entity can be generated. In this way, it is possible to generate a parameterized equivalent model entity when defining further parameter values ​​for the parameters of the parameterized equivalent model.

[0035] This invention, based on the concept of a specification of multiple entities in a parametric model, allows for greater automation through data-driven determination of parametric equivalent models. This is because parts of the parametric equivalent model can be automatically computed from multiple entities, such as at least one base entity, a parametric deformation map, or a decomposition of the parametric model into segments. Advantageously, in this process, the specified entities model the variability of the parametric model of the eyeglass frame elements as best as possible. If a single entity is used as the starting point, these steps require greater programming overhead, as the programmer must create routines to automatically identify individual eyeglass frame elements and modify them based on the parametric deformation map. By specifying multiple entities in the parametric model, alternatively, automated methods (e.g., machine learning methods) can be used to automatically generate parametric equivalent models based on the specified entities as automatically as possible.

[0036] Furthermore, this invention is based on the concept of using multiple entities when computing a parameterized equivalent model, which allows the quality of the parameterized equivalent model to be improved in the sense of the maximum possible similarity between the entities that the parameterized model and the parameterized equivalent model can generate, since recognition routines used to identify non-data-driven, i.e., individual features defined by programmers, are prone to errors.

[0037] Finally, multiple entities can be used to determine a parametric equivalent model that allows for changes to the entire object, rather than just a single detected geometric feature.

[0038] The second computer-implemented method according to the present invention for personalizing eyeglass frame elements by adapting a parametric model of the eyeglass frame element to the head of an eyeglass wearer includes the following method steps:

[0039] Specify multiple entities in the parameterized model;

[0040] A set of segments is determined for the parametric model of the eyeglass frame components;

[0041] Decompose the specified entity into segments of a set of segments;

[0042] For each segment in the group, a segment entity is generated using the entity of that segment selected from the specified decomposed entity;

[0043] Based on these segment entities, at least one basic segment entity and at least one parameterized deformation mapping of the at least one basic segment entity are determined.

[0044] In this case, the at least one parametric deformation mapping maps the at least one base segment entity to the segment entity of the parametric model. The parametric equivalent model is determined at least based on the group of segments and the at least one base segment entity of each segment in the group of segments and the at least one parametric deformation mapping.

[0045] In addition, biometric data related to the head of the eyeglass wearer is provided, and at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element is determined by optimizing a function that takes into account at least one surface point of at least one determined base segment entity of the parametric equivalent model of the eyeglass frame element and the provided biometric data related to the head of the eyeglass wearer.

[0046] In this case, a segment is a subset of the eyeglass frame elements, such as a portion of the front of the frame or a portion of the temple. If only one eyeglass frame element is available, such as in the form of a complete eyeglass frame, then the set of segments may include, for example, a bridge, temples, or connecting points.

[0047] The segment entity represents the entity of a segment in a parametric model or a parametric equivalent model.

[0048] The base segment entity represents the foundational entity of a segment in the parametric model or parametric equivalent model. This base segment entity is determined based on the selected segment entities of that segment.

[0049] The concept of the second method implemented by a computer according to the invention is based on the idea that a higher quality and more flexible parametric equivalent model can be obtained by: decomposing the parametric model of the eyeglass frame element into segments; and individually determining at least one basic segment entity and at least one parametric deformation mapping for each segment. Therefore, this allows the parametric equivalent model to fit particularly well to the characteristics of the corresponding segment, rather than mapping the deformation of at least one basic segment entity of the entire parametric model. This measure promotes greater variability and adaptability of the parametric equivalent model. Additionally, it also promotes a reduction in the complexity of at least one basic segment entity and the parametric deformation mapping, and thus a reduction in the complexity of the parametric equivalent model. Furthermore, the lower complexity of the parametric equivalent model simplifies the determination of parameter values ​​for the parameters of the parametric equivalent model, thereby saving computational time.

[0050] In this case, at least one base entity or base segment entity can be selected or calculated based on a specified entity or segment entity (e.g., by determining an average value).

[0051] If a parametric model of the eyeglass frame element is not yet available in the form of a segment or part, the second method implemented by a computer is particularly advantageous.

[0052] A third method for personalizing eyeglass frame components by adapting a parametric model of the frame components to the head of an eyeglass wearer includes:

[0053] A parametric equivalent model of a parametric model of an eyeglass frame element is determined by means of the following: determining a set of segments for the parametric model of the eyeglass frame element; determining a parametric segment model from the parametric model of the eyeglass frame element for each segment; determining the parametric equivalent model having at least one parameter as a segment equivalent model for each parametric segment model in the computer-implemented first method; and determining the parametric equivalent model based at least on the set of segments and the parametric segment equivalent model having at least one parameter.

[0054] Furthermore, biometric data related to the head of the eyeglass wearer is provided, and at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element is determined by performing a function that takes into account at least one surface point of the determined base entity of the segmental equivalent model of the parametric equivalent model of the eyeglass frame element and the provided biometric data related to the head of the eyeglass wearer.

[0055] In this context, the parametric segment model represents a parametric model that describes only one segment of the eyeglass frame element. This parametric segment model can be determined based on the parametric model of the eyeglass frame element, for example, if the eyeglass frame element has already been given in partial form.

[0056] The third computer-implemented method is based on a concept similar to the second computer-implemented method, specifically decomposing the parametric equivalent model into segments with the aforementioned advantages. While the second computer-implemented method generates entities of the parametric model of the eyeglass frame element, which are subsequently decomposed into segments, the third computer-implemented method decomposes the parametric model itself into segments, thus allowing for the determination of a parametric segment model for each segment. Then, a dedicated segment equivalent model is determined for each parametric segment model from the first computer-implemented method. This process is particularly advantageous if the parametric model of the eyeglass frame element is already available in the form of segments.

[0057] For example, eyeglass frame elements can be represented in a computer cell as a mesh or point cloud. Preferably, these objects can be represented as a mesh. Alternatively, the mesh can initially be generated via triangulation, for example, based on a given point cloud.

[0058] Specifically, the mesh is a triangular mesh consisting of surface points in the form of nodes, normal vectors at the nodes, and triangular surfaces. This continuous representation of the triangular mesh can be generated based on a shading algorithm.

[0059] Besides triangle-based mesh representations, there are other polygon-based or volume-based mesh representations, as explained in the July 11, 2019 Wikipedia article "Types of Meshes" (https: / / en.wikipedia.org / wiki / Types_of_mesh). For example, a two-dimensional mesh can be composed of triangular or quadrilateral elements. A three-dimensional mesh can be composed of pyramidal, cubic, or prismatic elements.

[0060] A physical entity of a parametric equivalent model of an eyeglass frame element can be generated based on a parametric equivalent model of the eyeglass frame element and a given set of parameter values ​​for the parameters of the parametric equivalent model. This parametric equivalent model contains a set of surface points in the form of 3D points on the surface of at least one eyeglass frame element.

[0061] According to the present invention, in a computer-implemented method for personalizing eyeglass frame elements by adapting a parameterized model of the eyeglass frame element to the head of an eyeglass wearer, a parameterized equivalent model of the parameterized model of the eyeglass frame element is determined, the parameterized equivalent model having at least one parameter.

[0062] In this process, biometric data related to the head of the eyeglass wearer are also determined. The biometric data related to the eyeglass wearer's head may consist of at least one surface point of a representation (e.g., a grid) of the eyeglass wearer's head. This data may be available in the computer unit, for example, in the form of surface points of the eyeglass wearer's head in a coordinate system. This can be achieved, for example, by recording the head from different recording orientations using an image processing device and calculating a 3D model of the head using a 3D reconstruction method or a SLAM method. To avoid having to calculate a complete 3D model, to minimize errors, and to save computation time, only a few 3D points of the head may be determined and adapted to a head model determined based on multiple exemplary data. For example, this head model may be determined using machine learning methods. Alternatively, the 3D model of the head may also be loaded into the computer unit from a storage medium or via a network. In this case, the eyeglass wearer's head is preferably represented as a grid in the computer unit. Alternatively or additionally, head dimensions (e.g., ear spacing, bridge of the nose width, or other length dimensions on the head) may also be determined as biometric data of the eyeglass wearer's head.

[0063] Then, in the second step, at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element is determined such that the entity of the parametric equivalent model of the eyeglass frame element generated based on the at least one parameter value fits the head as well as possible. The at least one parameter value is determined by optimizing a function that considers at least one surface point of the determined base entity of the parametric equivalent model of the eyeglass frame element and determined biometric data related to the wearer's head. In this case, the biometric data related to the wearer's head may be in the form of length and distance measures, or alternatively or additionally, in the form of surface points of the head, such as individual points like ear supports, or as a point cloud representing a part or the entire head. In this case, it is advantageous for the function to be optimized to also consider the parameters of the at least one parametric deformation map of the parametric equivalent model, which affects the relative positions of the surface points of the at least one base entity. The function to be optimized can minimize the distance between point clouds, such as the distance between a first point cloud consisting of at least one surface point of the base entity or base segment entity of the parametric equivalent model and a second point cloud consisting of at least one surface point of a representation (e.g., a mesh) of the eyeglass wearer's head. The function to be optimized can also consider specific points of the base entity of the parametric equivalent model of the eyeglass wearer's head and / or the eyeglass frame element, such as support points on the eyeglass frame element at the ears or nose, and corresponding support points on the eyeglass wearer's head. The eyeglass frame element can be adapted to the head by minimizing the distance between corresponding support points. The entity of the parametric equivalent model of the eyeglass frame element, generated based on at least one parameter value, is then adapted to the head. Alternatively, the function to be optimized can also adjust portions of the base entity of the parametric equivalent model of the eyeglass frame element based solely on biometric data of the head, such as temple length or bridge width. The function can also consider the eyeglass wearer's head size or the parametric equivalent model of the eyeglass frame element, such as bridge width, bridge width, lens size, or ear-point spacing. The function to be optimized may also include parameters relating to the relative positions of the underlying entities of the parametric equivalent model of the eyeglass frame elements, which are related to surface points on the head, such as rotation, translation, and scaling parameters. Alternatively, the parametric equivalent model may be adjusted by the user via entries in the user interface of the computer unit, based on biometric data related to the wearer's head. Further methods and details of adjustments in this regard are described in US2018 / 0336737 A1, EP 2 746 838 A1, and US 2016 / 0327811 A1, which are incorporated herein by reference and whose disclosures are incorporated into this specification.

[0064] A probability distribution or range of values ​​associated with the parameters of the parameterized model or its equivalent can be given or determined. In this case, the range of values ​​is a continuum of parameter values ​​bounded by a minimum and a maximum value. Alternatively, the range of values ​​can also be a set of discrete parameter values, for example, by sampling a continuous range of values ​​between the minimum and maximum values, such as at equal intervals. If a probability distribution is provided for the parameters, it is advantageous to choose parameter values ​​with higher probabilities. The range of values ​​or probability distribution of the model's parameters can also be determined based on a specified set of entities in the parameterized model or its equivalent.

[0065] In the current context, the probability distribution of parameters in a parametric model or parametric equivalent model is understood to describe the frequency of occurrence of each parameter value during the generation of different entities. If the probability distribution of a parameter's specific value is large, it means that the parameter's value is typical for many entities. In contrast, if the probability distribution of a parameter is small or approaches zero, it means that the parameter's value will not appear in a large number of entities.

[0066] Determining the range of parameter values ​​and / or probability distribution improves the manageability of the parameterized equivalent model for users, as parameter values ​​that generate unrealistic or undesirable entities are excluded in advance.

[0067] Entities generated based on the range or probability distribution of parameter values ​​in a parametric model can be implemented automatically, for example, by selecting the mean, median, or expected value. Alternatively, parameter values ​​can also be manually selected by the user through a user interface.

[0068] The parametric model of an eyeglass frame element can be adjusted using parametric mappings, which are applied, for example, to all surface points or a subset of surface points in the mesh of the eyeglass frame element. The eyeglass frame element can be modified by adjusting the values ​​of all or individual parameters of these mappings.

[0069] For a given entity in a parametric model or parametric equivalent model, changes in parameter values ​​will cause changes in points on the entity's surface.

[0070] A parametric equivalent model can contain the same parameters as the parametric model, or it can contain different parameters, additional parameters, or only a subset of the parameters of the parametric model.

[0071] A parameterized equivalent model can contain the following components, each of which can have the following parameters:

[0072] - This section;

[0073] - Number of sections;

[0074] - This is the at least one basic entity or basic segment entity that serves as the mesh;

[0075] - The at least one base entity or base segment entity is in the form of an index, which marks the selected base entity or base segment entity within the specified entity or segment entity;

[0076] - Allows for specific calculation rules to be determined based on specified entities or segment entities, for example by calculating the average value of specified (optionally normalized) entities or segment entities;

[0077] - This at least one parameterized deformation mapping;

[0078] - Additional features, such as ear support points, nose support points, and support curves at the ends of the temples; as approximate 3D lens planes for lenses to be fitted into the eyeglass frame; 3D boxes for approximating the front rims of the frame; nose pads.

[0079] - Post-processing routines;

[0080] - The range of parameter values ​​and / or probability distribution of the parameter values ​​in the parameterized equivalent model.

[0081] Therefore, the parametric equivalent model can be used as an alternative to the parametric model. Advantageously, the quality of the parametric equivalent model should be as high as possible. That is, for each entity in the parametric model, entities in the parametric equivalent model can be generated in a way that minimizes the deviation between the two entities.

[0082] The deviation between two entities can be determined based on their surface points. This deviation can be calculated as a standard based on a weighted sum, mean, maximum, or quantile of the distributions comprising the minimum deviations between the surfaces (e.g., surface points) of one entity and the surfaces of another entity. The deviation between the two entities can be determined, for example, by a one-sided or two-sided Hausdorff distance, as described in N. ASPERT, D. SANTA-CRUZ, T. EBRAHIMI, MESH: Measuring Errors between Surfaces using the Hausdorff Distance, Proceedings of the IEEE International Conference on Multimedia and Expo Lausanne, Switzerland (2002), pp. 1-4, which is cited herein and whose disclosure is included in this specification.

[0083] One-sided Hausdorff distance h and all minimum distances from one entity S to another entity S' The maximum value corresponds to, for example, the maximum Euclidean distance between a surface point of one entity S and the nearest corresponding surface point of another entity S':

[0084] .

[0085] In contrast, the two-sided Hausdorff distance H(S, S') describes the maximum value of the two one-sided Hausdorff distances between surfaces S and S':

[0086]

[0087] As an alternative to the deviation between surface points of an entity, the deviation between the surfaces themselves can also be determined, for example, based on mesh triangles or based on the skeleton of the entity determined by a skeletonization method.

[0088] To measure the quality of a parametric equivalent model of an eyeglass frame element, a quality criterion for the parametric equivalent model of the eyeglass frame element can be defined based on a group A of entities in the parametric equivalent model of the eyeglass frame element (e.g., specified entities not used to generate the parametric equivalent model and / or additional entities) and a group B of entities in the parametric equivalent model of the eyeglass frame element. In this case, for each entity in group A, group B includes entities generated based on the parametric equivalent model of the eyeglass frame element that have the minimum possible deviation between surfaces.

[0089] In this case, the quality standard can have a continuous range of values, for example, in the form of the maximum or average deviation between entities in group A and entities in group B.

[0090] Alternatively, binary value quality standards that are met or not met can be used. For example, such quality standards can be formulated in the form of conditions that the parametric equivalent model must satisfy to meet user quality requirements. For instance, maximum permissible deviations can be defined for different regions of a parametric model of an eyeglass frame element, where these deviations might occur between entities in group A and group B within a specified region. For example, a maximum permissible deviation of 0.05 mm can be defined for the surface in the nose support region, and a maximum deviation of 0.5 mm can be defined for the surface in the temple region.

[0091] This invention interprets the optimization of continuous quality standards as maximizing or minimizing the components (e.g., the at least one base entity, the group of segments, or the parametric deformation mapping) of the parametric equivalent model of at least one eyeglass frame element. This invention interprets the optimization of binary value quality standards as adjusting the parameters of the parametric equivalent model of at least one eyeglass frame element until a specified condition is met.

[0092] For each parameter of the parametric equivalent model of the at least one eyeglass frame element, a range of parameter values ​​and / or a probability distribution can be determined. To this end, multiple entities (e.g., a specified entity) can be represented based on the parametric equivalent model. Then, the range of parameter values ​​or the probability distribution of the parameters in the parametric equivalent model can be determined based on the parameter values ​​associated with the specified entity.

[0093] The equivalent model of at least one eyeglass frame element determined in the method described above specifically offers the following advantages:

[0094] Because of the high degree of automation in this method, the user incurs almost no overhead in generating a parametric equivalent model of the at least one eyeglass frame element. Specifically, machine learning methods that can largely automate the generation of parametric equivalent models based on specified entities can be used here.

[0095] Furthermore, this method requires very little computation time because its individual steps can be executed particularly efficiently. For example, the base entity and parametric deformation mapping can be determined simply by selection.

[0096] Furthermore, the generated parametric equivalent model of at least one eyeglass frame element is of particularly high quality, meaning that an entity generated based on a given parametric model of the eyeglass frame element can also be represented with small deviations based on the parametric equivalent model. The smaller the deviation, the higher the quality of the parametric equivalent model, as this makes it more suitable as a replacement for the parametric model.

[0097] Furthermore, the generated parametric equivalent model is less complex due to the defined structure of its underlying entities and parametric deformation mappings. This facilitates the particularly rapid adaptation of the parametric equivalent model to the eyeglass wearer's head, as the optimization of a low-complexity system requires low-complexity algorithms for optimization purposes and therefore less computation time. Moreover, the low complexity of the parametric equivalent model facilitates particularly rapid handling of parameter changes. This is because customers and opticians only accept a system for personalizing frames when the results of parameter changes during the adjustment process are immediately visible on the screen. However, in the case of complex parametric models, calculating new entities in the event of parameter changes typically requires several seconds of computation time.

[0098] When represented in the memory of a computer unit, the entity of the parametric equivalent model of at least one eyeglass frame element requires very little storage space. This is because what is stored is not the entire mesh, but the parameter values ​​of the elements of the parametric equivalent model. For example, if there is more than one base entity in the parametric equivalent model, the index of the selected base entity or the parameter values ​​of the at least one parametric deformation mapping is stored. Therefore, a database with many eyeglass frame models or eyeglass frame element models can be stored without much difficulty. At the same time, this measure reduces the transmission time between the adaptation system and the ordering system. Furthermore, the parametric equivalent model is therefore also suitable for compressing the entities of the parametric model. This is because the entities of the parametric model can be represented as entities of the parametric equivalent model of the parametric model, where only the parameter values ​​of the entities of the parametric equivalent model need to be stored instead of the entire mesh.

[0099] Parametric equivalent models offer users significant flexibility in generating entities. This is because entities can be selected not only from a specified, stored set of entities in the parametric models of eyeglass frame elements, but also for any parameter value (e.g., intermediate values). For example, given entities of parametric models with temple lengths of different lengths, entities of additional lengths can be generated based on the parametric equivalent model. Therefore, the parametric equivalent model can also fit the eyeglass wearer's head more accurately than a specified entity.

[0100] Because of these advantages, users can more comfortably work with the parametric equivalent models of the generated eyeglass frame elements and their generation methods.

[0101] The specified entities for the parametric model of an eyeglass frame component can be selected from a set of entities within the parametric model. This set can be generated by a modeling system that generates parametric models from this set. Based on this set of specified entities, individual methodological steps can be optimized to ensure a higher quality parametric equivalent model.

[0102] In this context, it is advantageous that the designated entity group comprises at least two entities of a parametric model of an eyeglass frame element generated based on different parameter values. Each designated entity represents a specific implementation of the parametric model for the selected set of parameter values. For example, the boundary of the range of parameter values, or its average or median, can be selected as the parameter value. Alternatively, the entities can also be determined by randomly selecting parameter values.

[0103] Let n be the number of parameters in the parameterized model. Then, within the corresponding parameter range, a total of k parameter values ​​are selected for each parameter. A set of specified entities is generated based on all combinations of these parameter values; this set therefore includes... k nEach entity. Advantageously, for example, k = 2 parameter values ​​are selected for each parameter based on a parameter value from the upper limit of the parameter value range and a parameter value from the lower limit of the value range. It is even more advantageous to have k = 5 parameter values ​​for each parameter. Alternatively, a different set of parameter values ​​can be selected for each parameter.

[0104] The designated entity is preferably located in a common coordinate system, such as the coordinate system of a parametric model of an eyeglass frame element. More preferably, the designated entity of the parametric model is positioned and oriented in the coordinate system such that the centroid of the corresponding entity corresponds to the center of the coordinate system. Alternatively, the plane of symmetry of the corresponding entity may contain one or both axes of the coordinate system.

[0105] For example, alignment can be calculated based on principal component analysis by determining the first two (orthogonal) principal components of the grid points and transforming all points of the grid in a way that maps the two principal components to coordinate axes (e.g., the first principal component is mapped to the applied axis and the second principal component is mapped to the ordinate axis).

[0106] The advantage of these alignment measures is that the parameterized equivalent model has the highest possible quality and its generation requires the least amount of computation time and can be performed in a highly automated manner.

[0107] The generation of specified entities can be automated using computer programs, thus saving users computation time and resources. Furthermore, this approach contributes to a high degree of automation in the method.

[0108] Advantageously, the specified entities of the parametric model are post-processed, at least in part, through algorithms used to correct errors and / or improve the visual impression on eyeglass wearers and / or for smoothing. Errors can be, for example, topological defects such as holes or irregular triangulation measurements, such as the irregular density or size of triangles on a mesh surface. A higher quality parametric equivalent model based on the specified entities can be obtained based on this preprocessing step.

[0109] Elements of a parametric equivalent model can be determined based on a specified entity. This determination can be performed manually by a programmer or user, or automatically based on machine learning methods. To determine, for example, the parametric equivalent model of the at least one base entity, the at least one parametric deformation map, or the group of segments, it is advantageous to optimize the criteria based on a weighted sum, mean, maximum, and quantile of the distribution of deviations between the surfaces (e.g., surface points) of the specified entity including the parametric model and the surfaces of all those entities whose parametric equivalent models can be generated based on specific parameter values.

[0110] Furthermore, it is advantageous to apply methods for identifying inflection points in the signal and / or mesh segmentation methods and / or multivariate fitting methods and / or skeletonization methods and / or machine learning methods during the decomposition of the parameterized model of the eyeglass frame element into segments of the group of segments.

[0111] To automatically decompose the solid of a parametric model of an eyeglass frame element into segments within a set of segments based on a method for detecting inflection points, surface points of the solid along spatial axes are projected onto a plane. Based on the projected points, an algorithm selects a subset whose preimage is associated with the contour of the eyeglass frame element. This subset of projected points can be understood as a sequence of discrete sampled values ​​of a signal. The inflection points of this signal can then be determined algorithmically. These inflection points are ultimately used to determine the boundaries of the segments within the set of segments.

[0112] The advantage of this process is that entity decomposition can be performed in a fully automated manner. This is because the algorithms used to detect inflection points do not require semantic information about the properties of individual segments or their boundaries, or about how to find these in the data. This significantly reduces user overhead.

[0113] Alternatively, machine learning methods can be used to automatically decompose an entity into segments within that group of segments.

[0114] Alternatively, other grid segmentation methods can be used, such as those described in the article " A Survey on Mesh Segmentation Techniques [Overview of Mesh Segmentation Technology] Ariel Shamir, Computer Graphics Forum [Computer Graphics Forum], Volume 27, Issue 6, 2008, pp. 1839-1856 The method described in the document can also be used, such as the multivariate adjustment method described in the document. Using Multivariate Statistics [Using multivariate statistics], Barbara G. Tabachnick, Linda S. Fidell, Jodie B. Ullman, Pearson Verlag. 2007 As presented in the book.

[0115] As an alternative, as in the article " Skeleton Extraction by Mesh Contraction [Extracting the skeleton through mesh shrinkage], Oscar Kin-Chung Au,Chiew-Lan Tai, Hung-Kuo Chu, Daniel Cohen- Or, Tong-Yee Lee, Proceedings of SIGGRAPH 2008 The skeletonization method described in the "SIGGRAPH 2008 Proceedings" can also be used for segment entities.

[0116] This invention is based on a comprehensive review of the aforementioned books and two articles, the disclosures of which are included in the specification of this invention.

[0117] Based on skeletonization methods, for example, the mesh of a solid can be generated into a low-complexity structure in the form of a skeleton. In this case, the skeleton of a 3D object includes all the interior points of the object, which are the center points of the largest sphere contained within the object.

[0118] Each surface point of the solid can then be assigned to the nearest point in the generated skeleton. Now, instead of the solid's mesh, the less complex skeleton can be decomposed into regions. All surface points associated with the skeleton regions then form segments. This process saves computation time.

[0119] Another advantage is that the generated skeleton can also be used in subsequent method steps to determine the parametric deformation mapping. This is because mapping one entity to another based on the associated skeleton also saves complexity and computation time.

[0120] The at least one parametric deformation mapping is used to map a base entity or base segment entity to another entity or corresponding segment entity of the parametric model of the at least one eyeglass frame element. In this case, the parametric deformation mapping is defined in the form of a function with parameters to be determined.

[0121] For example, an affine map describing the rotation, translation, and scaling of a segment can be selected as a parametric deformation map.

[0122] Advantageously, the parametric deformation mapping of the parametric equivalent model originates from a group including affine mappings, polynomials, polynomial surfaces, Bézier curves, splines, or NURBS. This enables higher quality parametric equivalent models and shorter computation time during the representation or fitting of eyeglass frame elements to the head.

[0123] Furthermore, the automation level of this method can be improved, for example, by selecting a low-complexity parametric deformation mapping with few parameters (e.g., an affine mapping). This is because, in this case, the parameter values ​​of the parametric deformation mapping can be automatically determined based on an algorithm used to minimize the deviation between the entities of the parametric model and the parametric equivalent model.

[0124] Furthermore, machine learning methods can be used to determine the components of the parametric equivalent model, specifically the at least one base entity and the at least one parametric deformation mapping. Preferably, principal component analysis can be used in this case. For this purpose, the designated entities are represented by point clouds or voxel rasters. These can be represented as vectors, which, for example, contain the coordinates of points or information about whether each voxel is located inside or outside the eyeglass frame element. The vectorized designated entities of the parametric model can be used to first determine the average value of the designated entities. This then forms the base entity. The average value can be subtracted from each designated entity, and the covariance matrix of the entity can be calculated from this. Its diagonalization allows the determination of the eigenvectors and eigenvalues ​​of the covariance matrix. To achieve lower complexity of the parametric equivalent model, eigenvectors can be selected only for large eigenvalues. The designated entities of the parametric model or its segments and the additional entity I can now be represented by the parametric deformation mapping as the base entity b with the average value and n eigenvectors. vi The linear combination of the forms is approximately represented as follows:

[0125]

[0126] Then, the parameterized equivalent model consists of the basic entity b and the feature vector. v i The composition is then determined in order to represent a specific entity based on a parametric equivalent model, and the parameter values ​​of the parametric deformation mapping are then determined. Neural networks can also be used to automatically calculate underlying entities and deformation mappings.

[0127] Preferably, the parameterized equivalent model is stored in the memory of the computer unit.

[0128] The advantage of this method is that it is applicable to parametric models of at least one eyeglass frame element with any type of surface. The surface type is defined as the maximum possible number of cuts along a non-intersecting closed simple curve, such that the surface remains continuous after all cuts. Therefore, the surface type represents the number of holes on the surface. This improves the user experience of the method, as it is not limited to a single type of eyeglass frame element.

[0129] Advantageously, segments in this group of segments of the parametric equivalent model are labeled as static, movable, or deformable.

[0130] Labeling can be performed automatically based on a clustering method that analyzes continuous surface points based on the movement of various entities (e.g., specified entities) through a parameterized model of the eyeglass frame elements.

[0131] Based on this label, the quality of the parametric equivalent model can be improved because the parametric deformation mapping can be appropriately selected based on the movement of the corresponding segments.

[0132] When selecting segment entities from a specified set of entities decomposed into segments, one segment entity is sufficient for segments marked as static. For segments marked as movable or deformable, the presence of at least two segment entities, preferably five, is beneficial for the accuracy of the parametric equivalent model.

[0133] Furthermore, it is particularly advantageous that the parametric deformation mapping of segments marked as static is a linear mapping and / or the parametric deformation mapping of segments marked as movable is an affine mapping and / or the parametric deformation mapping of segments marked as deformable is based on polynomial, polynomial surface, Bézier curve, spline or NURBS approximation.

[0134] Therefore, the complexity of parametric deformation mapping is reduced by adapting to the movement of the segments. This saves computation time and improves the quality of the parametric equivalent model.

[0135] Preferably, when the parameters of the parametric equivalent model of the eyeglass frame element change, the triangular structure of the mesh (i.e., the topology and links of the triangles) remains unchanged instead of being recalculated. This eliminates the time-consuming step of triangulating surface points to adjust the triangular mesh. This saves computation time and simultaneously produces a parametric equivalent model of the eyeglass frame element with lower complexity.

[0136] An advantageous development of the present invention provides methodological steps for determining the parameterized equivalent model to be iterated. This measure ensures higher quality parameterized equivalent models because the various elements of the parameterized equivalent model are interdependent and can be better optimized in this way.

[0137] Advantageously, the segments in this group are arranged hierarchically in a tree structure, such that the nodes connected in the tree structure are associated with segments in the parametric model that share a common cutting edge or cutting surface. Thus, the interconnecting nodes of the tree indicate the spatial neighborhood of the associated segments.

[0138] Furthermore, advantageously, each segment in the tree structure is positioned and oriented relative to its parent segment in a coordinate system. In this case, the segment can contain a dedicated local coordinate system and additionally its position and orientation relative to its superior segments in the tree structure. Due to the relative orientation of the segments with respect to each other, this generally produces a composition of rigid body transformations used to adjust the underlying entity of the eyeglass frame element. For example, rigid body transformations can be encoded as kinematic chains, as described in the Wikipedia article "Forward Kinematics" dated June 28, 2019.

[0139] The hierarchical arrangement of segments simplifies the calculation of parameter values ​​in the parametric equivalent model, as these parameter values ​​can be determined incrementally for each node along the hierarchical structure and can be determined based on the calculated parameter values ​​of the parent nodes. This saves computation time and improves the quality of the parametric equivalent model.

[0140] If there are at least two eyeglass frame elements, these eyeglass frame elements can be arranged additionally or as alternatives to segments in the hierarchical tree structure.

[0141] Since the parameter values ​​are determined independently of other segments for each segment of at least one basic entity of the at least one eyeglass frame element, discontinuities may exist at segment boundaries. To improve the quality of the parametric equivalent model and enhance the visual impression on the eyeglass wearer, the entities of the parametric equivalent model can be post-processed based on an algorithm designed to avoid discontinuities at segment boundaries. This measure can be provided in additional method steps.

[0142] The post-processing steps of the parametric equivalent model of at least one eyeglass frame element may also include algorithms for correcting errors and / or for improving the visual impression of the eyeglass wearer and / or for smoothing the mesh.

[0143] For example, a smoothing method can be chosen as a post-processing step. The type of the post-processing method and its parameters can be determined and can be stored separately in the parameterized equivalent model.

[0144] Furthermore, it is advantageous to integrate symmetry assumptions related to individual segments (such as the left and right temples) into the parametric equivalent model. For example, to create a parametric equivalent model of the entire eyeglass frame, its symmetry means that it is sufficient to have a parametric equivalent model for only the left or right temple. The entity of the corresponding other temple can be determined by reflections in the plane of symmetry of the eyeglass frame and alignment of the front of the frame. This measure can save computation time, memory space, and transfer time.

[0145] When the parametric model of the eyeglass frame element varies considerably, the complexity of the parametric equivalent model of at least one eyeglass frame element can also be reduced to save computation time and optionally reduce user costs. To this end, a relatively large set of basic entities for the parametric equivalent model of the eyeglass frame element is selected in a way that best represents the range of variation of the eyeglass frame element.

[0146] Therefore, eyeglass frame elements with small variations (such as temples that typically only vary in overall length) can be directly selected from a set of basic entities (e.g., a set of basic entities for different temple lengths). In this way, it is not necessary to determine a parametric deformation map based on an algorithm and calculate its parameter values.

[0147] Advantageous embodiments of the invention also provide additional features to be calculated for the parametric equivalent model of the eyeglass frame elements. These additional features come from a group including ear support points, nose support points, support curves at the temple ends, 3D lens planes, 3D boxes, and nose pads. These additional features make it easier to adapt the parametric equivalent model of the eyeglass frame elements to the eyeglass wearer's head based on certain orientation points detected on a head model of the eyeglass wearer. This improves the manageability of the parametric equivalent model for the user.

[0148] In this invention, data format is understood to mean a representation of information or data that can be processed by a computer unit, specifically that can be read into and stored in the computer unit, such as a file in the hard disk space of the computer unit.

[0149] Advantageously, the parametric equivalent model is provided in a data format different from that of the parametric model, specifically in a data format independent of the system that generated the parametric model. For example, if the parametric model can be used as a CAD model, the parametric equivalent model can be provided in a data format suitable for the specific system in which the parametric equivalent model should be used, such as an adaptation system for fitting eyeglass frame components to the head of an eyeglass wearer.

[0150] Determining a parametric equivalent model with a format-independent representation is simplified by specifying the entities of the parametric model. The parametric model is then available in a specific format, such as the format of the modeling program used by the designer. However, the specified entities can be presented as a mesh. Therefore, these specified entities are format-independent of the modeling program and can be stored in different data formats.

[0151] Advantageous developments provide a parametric equivalent model of the parametric model of an eyeglass frame element used in a computer-implemented method for representing and / or compressing a parametric model of an eyeglass frame element in a computer unit, the parametric equivalent model having at least one parameter.

[0152] In this scenario, in the first step, the corresponding parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element is determined by optimizing a criterion based on a weighted sum, mean, maximum, and quantile of the deviation distribution between the surface (e.g., surface points) of a given entity including the parametric model and the surface (e.g., surface points) of the entity in the parametric equivalent model generated based on the at least one parameter value. The at least one determined parameter value can then be stored in the memory of the computer unit.

[0153] The advantage of this method is that, if a parametric equivalent model of the at least one eyeglass frame element is available, a given entity of the parametric model of the at least one eyeglass frame element can be represented based on a very small number of parameter values. Therefore, entities can be stored in a very memory-efficient manner. This allows for a significant reduction in memory requirements, particularly when the entity groups in the eyeglass frame database are relatively large. Due to the smaller data volume, this can also be accompanied by a significant reduction in the transmission time of adapted eyeglass frame elements, for example, between the optician's fitting system and the ordering system or the eyeglass wearer's personal computer unit.

[0154] In the method for personalizing a parametric model of at least one eyeglass frame element and / or in the method for representing and / or compressing an entity of a parametric model of an eyeglass frame element, and when passing an entity of a parametric model of an eyeglass frame element, it is advantageous to minimize the distance between point clouds when optimizing at least one parameter of the parametric equivalent model.

[0155] In methods for personalizing parametric models of eyeglass frame components, point clouds are provided as surface points of the base entity of the parametric equivalent model and surface points of the mesh of the eyeglass wearer's head. In this case, for example, the distance between the ear support point on the temple and the surface point on the eyeglass wearer's ear is minimized.

[0156] In a method for representing and / or compressing a parametric model of an eyeglass frame element, point clouds are given as surface points of a selected base entity of the parametric equivalent model of the eyeglass frame element, and as surface points of the entity to be represented and / or compressed. In this case, Hausdorff distance is used, for example, to minimize the deviation of all surface points of the two entities.

[0157] Methods for minimizing the distance between point clouds include, for example, the Iterative Closest Point (ICP) algorithm, which is discussed in the paper "..." Efficient Variants of the ICP Algorithm [Efficient variant of the ICP algorithm] Szymon Rusinkiewicz, Marc Levoy, Proceedings of the 3DIM Conference [3DIM Conference Proceedings], Quebec, 2001, pp. 145-182 The text is described in conjunction with various variations, the entire contents of which are incorporated herein by reference, and the disclosures of which are included in the specification of this invention.

[0158] The advantage of this algorithm is that it can determine the parameter values ​​of the parameterized equivalent model with particularly high accuracy while minimizing overhead and computation time. This improves the quality and manageability of the parameterized equivalent model.

[0159] The computer program product according to the present invention comprises a computer program having program code for performing the above-described method steps when the computer program is loaded into a computer unit and / or executed on the computer unit.

[0160] An apparatus for personalizing and adapting a parametric model of an eyeglass frame element to the head of an eyeglass wearer includes a computer unit loaded with a computer-implemented method for adapting the parametric model of the eyeglass frame element to a head representation in a coordinate system.

[0161] An apparatus for representing and / or compressing a parametric model of a given entity of an eyeglass frame element includes a computer unit having a memory in which a computer-implemented method for representing and / or compressing the given entity in the memory of the computer unit is loaded.

[0162] The system according to the invention having a device for performing the following operations uses at least one of the parameter values ​​determined by the parameterized equivalent model: producing personalized eyeglass frame elements in the method for personalizing eyeglass frame elements described above, or grinding eyeglass lenses into eyeglass frame elements as personalized in the method for personalizing eyeglass frame elements described above. Attached Figure Description

[0163] Advantageous exemplary embodiments of the present invention, schematically depicted in the accompanying drawings, are described below.

[0164] In detail:

[0165] Figure 1 The parametric model of the eyeglass frame element is shown in the form of a CAD model of an eyeglass frame with different additional frame elements;

[0166] Figure 2 The grid of an eyeglass frame element with surface dots and a triangular mesh is shown;

[0167] Figure 3 A method for personalizing eyeglass frame elements by adapting a parametric model of the frame elements to the head of an eyeglass wearer is shown.

[0168] Figure 4 A method for determining a parametric equivalent model of an eyeglass frame element in the form of a temple is shown;

[0169] Figure 5 An alternative method is shown for determining a parametric equivalent model of an eyeglass frame element in the form of the front of the frame;

[0170] Figure 6 An alternative method is shown for determining a parametric equivalent model of an eyeglass frame element in the form of the front of the frame;

[0171] Figure 7 A coordinate system is shown for arranging entities based on their centroid and symmetry plane;

[0172] Figure 8 The solid model of the front part and temples of the eyeglass frame is shown;

[0173] Figure 9 The basic entity of the CAD model for determining the front of the eyeglass frame is shown;

[0174] Figure 10 This illustrates the decomposition of a solid in a CAD model into segments of a set of segments;

[0175] Figure 11A parametric equivalent model of a CAD model for determining the front of a glasses frame based on a specified entity using principal component analysis is shown. This parametric equivalent model has a base entity and a parametric deformation mapping.

[0176] Figure 12 The method steps for determining a parametric equivalent model of an eyeglass frame element based on a specified entity are shown;

[0177] Figure 13 The parametric equivalent model of an eyeglass frame element shows the arrangement of segments in a hierarchical tree structure.

[0178] Figure 14 A method for personalizing eyeglass frame components is shown;

[0179] Figure 15 A method is shown for representing a physical entity of a parametric model for compressing eyeglass frame elements;

[0180] Figure 16 The projection point is shown by projecting the surface points of the grid at the front of the frame onto a plane.

[0181] Figure 17 The solid upper and lower lens rings of the CAD model of the front of the eyeglass frame are shown;

[0182] Figure 18 The signal is shown as consisting of partial signals and inflection points;

[0183] Figure 19A , Figure 19B and Figure 19C The calculated inflection points and average values ​​of partial signals for the upper and lower spectacle lens rims via projected surface points are shown.

[0184] Figure 20A , Figure 20B The parametric equivalent model of the front of the eyeglass frame is shown as two entities decomposed based on the inflection point determined in the signal;

[0185] Figure 21 This shows the solid decomposition of the CAD model of the temple into two segments;

[0186] Figure 22 This demonstrates how to optimize the decomposition of an entity into segments by changing the parameter values ​​of the segments;

[0187] Figure 23A , Figure 23B and Figure 23C The parameter values ​​of the parametric deformation mapping of the basic segment entity and the corresponding segments of other entities in the CAD model of the temple are shown based on the ICP algorithm.

[0188] Figure 24A , Figure 24B and Figure 24C The parameter values ​​of the parametric deformation mapping of the base segment entity and the corresponding segments of other entities in the CAD model that determines the front of the eyeglass frame are shown.

[0189] Figure 25 The method steps for generating a mesh based on a parametric equivalent model and given parameter values ​​are shown; and

[0190] Figure 26A , Figure 26B and Figure 26C The solid model of the parameterized equivalent model of the connecting element is smoothed based on a post-processing step for smoothing at the segment boundaries. Detailed Implementation

[0191] Figure 1 A parametric model of the eyeglass frame element 24 is shown in the form of a CAD model 22 of the eyeglass frame, which has various other eyeglass frame elements 24, especially the front of the frame, temples and connecting elements.

[0192] If these eyeglass frame components 24 are already marked in the CAD model, the eyeglass frame component 24 for which a parametric equivalent model should be determined can be directly selected. If individual eyeglass frame components 24 are not marked, or if this is not desired, the parametric equivalent model of the entire eyeglass frame can be determined.

[0193] This method does not require the availability of a parametric model from the frame manufacturer itself—a set of 30 entities is sufficient.

[0194] The entity 30 of CAD model 22 can preferably be used as mesh 26. Figure 2 A grid 26 of the eyeglass frame element 24 is shown. The surface of the grid 26 consists of triangles defined based on surface points 28 in the form of points on the surface of the eyeglass frame element 24. Entities 30 can be stored, for example, as the grid 26 in the database 42.

[0195] Figure 3 Method steps 10, 10', 10" are shown for personalizing an eyeglass frame element by adapting a parameterized model of the frame element to the head of an eyeglass wearer. A parameterized model of the eyeglass frame element 24 is given in the first method step 2. For this parameterized model, in further method steps 4, 4', 4" a parameterized equivalent model of the eyeglass frame element 24 is determined, the parameterized equivalent model having at least one parameter. In this case, the parameterized equivalent model can be determined in three different ways, the method steps of which are described in... Figure 4 , Figure 5 and Figure 6As described in the figure. In another method step 6, biometric data 31 related to the head of the eyeglass wearer is provided (e.g., determined). Finally, in the final method step 8, at least one parameter of the parameterized equivalent model is determined by optimizing a function used to adapt the parameterized equivalent model to the head of the eyeglass wearer.

[0196] Figure 4 Method steps of method 4 for determining a parametric equivalent model of eyeglass frame element 24 are shown, wherein for a given parametric model of eyeglass frame element 24, the parametric equivalent model has at least one parameter.

[0197] Figure 4 The eyeglass frame element 24 shown is a temple. This temple can be presented as a parametric model in the form of CAD model 22. However, method 4 can also be applied to the parametric model of the entire eyeglass frame.

[0198] In the first method step 12 of method 4, entities 30 of a parametric model of the eyeglass frame element 24 in the form of temples are specified. Based on these specified entities, at least one base entity 38 is determined in the second step 14, and at least one parametric deformation mapping f(b, α) is determined in the third step 16. This at least one parametric deformation mapping maps the base entity b to the entity 30 of the parametric equivalent model based on parameters in the form of a parameter vector α. Various entities 30 of the parametric equivalent model can be generated by inserting different parameter values ​​of α; for example, the result is that the length and / or width of the temples can be changed so that the eyeglass frame element 24 can be adapted to the head of the eyeglass wearer.

[0199] The steps of method 4 used to generate the parametric equivalent model of eyeglass frame element 24 can be repeated in multiple iterations 18.

[0200] Figure 5 The method steps of an alternative method 4 for determining a parametric equivalent model of eyeglass frame element 24 are shown, wherein for a given parametric model of eyeglass frame element 24, the parametric equivalent model has at least one parameter.

[0201] Figure 5 The eyeglass frame element 24 shown is the front part of the frame. This front part of the frame can be presented as a parametric model in the form of CAD model 22.

[0202] In the first method step 12 of method 4', multiple entities 30 of the parametric model of the eyeglass frame element 24 in the form of the front of the frame are specified. In the second step 13, a set of segments 40 of the parametric model of the eyeglass frame element 24 are determined. In a further step 15, the specified entities 30 of the parametric model are decomposed into segments 40 within the set of segments 40. In the next step 17, a set of segment entities 43 are selected from the decomposed specified entities. Therefore, for each segment 40 in the set of segments 40, the corresponding segment 40 is selected from the decomposed specified entities 30 and the selected segments 40 are combined to form a set of specified segment entities 43, for example... Figure 5 The upper left portion of the front of the middle frame is shown. In another method step 20, a basic segment entity 39 is determined for each segment 40. Therefore, as in method 10 described above, the basic entity 38 and the basic segment entity 39 are determined based on the specified segment entities. Additionally, in another method step 21, for each basic segment entity b... i Each segment i determines the parameterized deformation mapping f i (b i , Deformation mapping is based on parameters. The basic segment entity b i The other segment entity 43 of segment i is mapped onto the parameterized equivalent model of the eyeglass frame element 24.

[0203] In this case, the specified entity 30 can be generated by changing the parameter values ​​of the parameters of the CAD model 22.

[0204] In this case, it is advantageous that the specified entity 30 is available in a single coordinate system 32, such as Figure 7 As shown. Furthermore, it is advantageous to position and orient the specified entity 30 in coordinate system 32 in such a way that the centroid 36 of the corresponding entity 30 corresponds to the center of coordinate system 32 and / or the plane of symmetry 34 of the corresponding entity 30 contains the axis of coordinate system 32.

[0205] Furthermore, the specified entity 30 can be preprocessed in preprocessing step 44 to correct errors, such as topological defects like holes or irregular triangulation measurements, such as the irregular density or size of surface triangles, and / or to improve the visual impression of entity 30 on the eyeglass wearer. For this purpose, a Poisson surface reconstruction algorithm can be used, for example, as described in the article "..." Poisson Surface Reconstruction [Poisson Surface Reconstruction] Michael Kazhdan, Matthew Bolitho and Hugues Hoppe, Proceedings of the fourth Eurographics symposium on Geometry processing [Proceedings of the 4th European Symposium on Geometric Processing in Computer Graphics] 2006 The entire contents of the article described herein are cited and the disclosures of the article are included in the specification of this invention.

[0206] The steps of method 4' used to generate the parametric equivalent model of eyeglass frame element 24 can be repeated in multiple iterations 18.

[0207] Figure 6 The method steps of an alternative method 4” for determining a parametric equivalent model of eyeglass frame element 24 are shown, wherein for a given parametric model of eyeglass frame element 24, the parametric equivalent model has at least one parameter.

[0208] Figure 6 The eyeglass frame element 24 shown is the front part of the frame. This front part of the frame can be presented as a parametric model in the form of CAD model 22.

[0209] In step 12 of the first method of method 4", a set of segments 40 of the parametric model of the eyeglass frame element 24 is determined. Furthermore, a parametric segment model is determined for each segment 40 based on the parametric model. For this purpose, the parametric model can be in the form of individual segments, for example, available in a CAD file containing multiple parts of the eyeglass frame. Then, in another step 19, the parametric model is determined based on the above... Figure 4 The interpretation method determines the equivalent model of each parameterized segment for each parameterized segment model. The parameterized equivalent model of the eyeglass frame element 24 then includes the parameters of the group of segments and the equivalent models of each segment.

[0210] When determining the elements of the parametric equivalent model, it is advantageous to optimize the criteria in each case based on a weighted sum, average, maximum, and quantile of the distribution of deviations between the surface of the specified entity 30, including the parametric model, and the surface of all those entities 30 that can be generated based on specific parameter values ​​of the parametric equivalent model of the at least one eyeglass frame element 24.

[0211] Another advantage is that it provides a parametric equivalent model in a different data format than the parametric model. This is because it allows the use of the parametric equivalent model to be independent of the program and the data format available for the parametric model.

[0212] In the second step 13 of method 10', a set of segments 40 of the parametric model of the eyeglass frame element 24 is determined. This measure aims to reproduce as closely as possible the virtual way in which the eyeglass frame manufacturer produces the parametric model.

[0213] To determine the group of segments 40, the influence of various parameters (such as frame size, temple length, bridge width, tilt angle, and flare angle) of the CAD model 22 created by the frame manufacturer on the geometry of the parametric model of the frame element 24 can be examined based on a designated entity 30 of the parametric model of the frame element 24. Figure 8As shown. For example, the entity 30 can then be analyzed based on the movement of the surface points 28 of the grid 26 on various entities 30. For example, all surface points 28 that follow the same movement or all surface points 28 that do not move can be combined to form a segment 40.

[0214] like Figure 8 As shown in Figure A, the dimensions of the eyeglass frame are scaled by a grid of 26 in all spatial directions. Figure 8 In B, the width of the nose bridge is scaled horizontally across the frame. Figure 8 The tilt angle in C causes the area at the front of the frame where the temples are mounted to move vertically, while Figure 8 The angle in D causes these regions to move horizontally. Figure 8 The temple length in E scales the length of the temple. For example, for the front of the frame, this group of segments 40 may contain twelve elements.

[0215] For segment 40 of the set of segments 40 of the parametric equivalent model of at least one eyeglass frame element 24, additional features for adapting the parametric equivalent model to the head of the eyeglass wearer can be determined, such as ear support points on the temples, support curves at the ends of the temples, a 3D lens plane approximating the lens to be adapted into the eyeglass frame, a 3D box approximating the front rim of the frame, and nose pads and / or nose support points at the front of the frame. This additional data may require manual user interaction, such as by selecting points or lines in the data displayed on the screen.

[0216] In step 15 of method 10', the designated entity 30 at the front of the frame is decomposed into segments 40 within the group of segments 40. For example, based on... Figures 18 to 21 As described, entity 30 can be manually segmented by the user through input via a user interface or automatically segmented by an algorithm.

[0217] For example, the front part of the frame is broken down into Figure 10 The twelve segments shown are labeled with numbers 1 to 12. The drawn planes each indicate the segment boundary 41. These planes can be determined, for example, based on an algorithm used to detect inflection points 74, as described further below.

[0218] If the designated entity 30 can be used as a mesh 26, it is advantageous to divide these designated entities into non-intersecting segments 40 within the group of segments 40, in such a way that each segment entity 43 consists of a continuous set of surface points 28 associated with the triangulation of the mesh 26. In this case, the triangulation of the mesh 26 remains unchanged even after the designated entity 30 is decomposed into segments 40.

[0219] In the second step 14 of method 10, at least one base entity 38 is provided by the specified entity 30 based on the parameterized model of the eyeglass frame element 24, such as Figure 9 As shown. Specifically, entity 30, generated based on the average or median of the range of values ​​of the probability distribution of the corresponding parameters or the expected value of the probability distribution, is appropriate here.

[0220] Alternatively, one of the specified entities 30 can be selected as the base entity 38. In this case, the at least one base entity 38 can be selected in a way that the other entities 30 (e.g., the remaining specified entities 30) can be reproduced with minimal possible error by applying a parametric deformation mapping.

[0221] Alternatively, the at least one base entity 38 can be selected based on user input via a user interface or automatically through an algorithm. In this case, the algorithm can evaluate quality criteria, such as the deviation between the entity 30 reproduced based on the parametric equivalent model and the specified entity 30.

[0222] The at least one base segment entity 39, i.e., the base segment entity 39, can be determined in the same manner based on the designated entity 30 that has been decomposed into segments 40.

[0223] In the final step 16 of method 10, at least one parametric deformation mapping is determined for the at least one base entity 38 to map the at least one base entity onto another entity 30 of the parametric model of the eyeglass frame element 24. In this case, the at least one parametric deformation mapping is defined in the form of a mapping f(b, α) having a parameter α that is to be determined to change the base entity b.

[0224] For example, an affine map describing the rotation, translation, and scaling of segment 40 can be selected as the deformation map.

[0225] Advantageously, the at least one parametric deformation mapping of the parametric equivalent model is derived from a group including affine mappings, polynomials, polynomial surfaces, Bézier curves, splines, or NURBS.

[0226] Specifically, advantageously, segments 40 in the group of segments 40 of the parametric equivalent model are labeled as static, movable, or deformable.

[0227] Of particular advantage is that the parametric deformation mapping of the segment 40 marked as static is a linear mapping, the parametric deformation mapping of the segment 40 marked as movable is an affine mapping, and the parametric deformation mapping of the segment 40 marked as deformable is based on a polynomial, polynomial surface, Bézier curve, spline, or NURBS approximation.

[0228] The segments 40 marked as movable or deformable and not following uniform movement can be the connecting surfaces between the eyeglass frame element 24 and / or the segments 40. These connecting surfaces include contact curves in the corresponding contact areas with adjacent segments 40. It may be advantageous for these connecting surfaces to define additional connection conditions in the form of points and normal vectors at several points of the contact curves.

[0229] For each segment 40 in the group of segments 40, at least one base segment entity 39 and at least one parameterized deformation map are determined in such a way that the at least one parameterized deformation map maps the base segment entity 39 to the other segment entity 43 with the smallest possible deviation.

[0230] Advantageously, the elements of the parametric equivalent model of the eyeglass frame element 24 are determined using an algorithm that minimizes the deviation between the entity 30 of the parametric model and all the generative entities 30 of the parametric equivalent model, specifically the group of segments 40, the at least one basic entity 38, and / or the parametric deformation mapping.

[0231] Machine learning methods can be used to determine the elements of the parameterized equivalent model, specifically the at least one basic entity 38 and the at least one parameterized deformation map. This is equally applicable to the determination of the at least one basic segment entity 39 and the at least one parameterized deformation map in method 10'.

[0232] Preferably, principal component analysis can be used here, such as... Figure 11 The depiction. Then, the average value of entity 30 is used to form the base entity b. After subtracting the average value, the eigenvectors based on the covariance matrix of entity 30 are obtained. v i Determine the parameterized deformation mapping. To achieve lower complexity of the parameterized equivalent model, we can select only n eigenvectors for the n largest eigenvalues.

[0233]

[0234] If a specific entity 30 of the CAD model 22 of the eyeglass frame element 24 is available, that specific entity can be represented for the CAD model 22 of the eyeglass frame element 24 based on the parametric equivalent model of the at least one eyeglass frame element 24 as follows. First, the entity 30 is decomposed into segments 40 of the group of segments 40 in the parametric equivalent model of the at least one eyeglass frame element 24. Then, the base entity 38 of the parametric equivalent model of the eyeglass frame element 24 is selected. Then, a specific deformation mapping for each segment 40 can be calculated, which maps the corresponding segment 40 of the base entity 38 to the corresponding segment 40 of the specific entity 30, for example, based on the following... Figure 18 and Figure 21Further description. Therefore, entity 30 can be approximately represented based on the parametric equivalent model of the eyeglass frame element 24 simply by specifying the parameter values ​​of the parametric deformation mapping of the selected base entity 38 and each segment 40 of the selected base entity 38.

[0235] Figure 12 This illustrates how to determine the parametric equivalent model of the eyeglass frame element 24 for a specified entity 30 based on a common parametric model. In this case, the specified entity 30 can be stored in the eyeglass frame manufacturer's database 42 in the form of a mesh 26.

[0236] To fix visual or topological defects, the specified entity can be preprocessed in preprocessing step 44.

[0237] Based on Figure 8 As described, in the next step, a set of suitable segments 40 is determined for each eyeglass frame element 24 by identifying the relevant frame parameters from the manufacturer. In this case, the parametric model of the eyeglass frame element may have already been presented in a segmented form, that is, subdivided into segments. Figure 6 Method 4 described in the text can be used to generate a parameterized equivalent model by means of determining the parameterized equivalent model of each segment.

[0238] If there is no available segmentation of the parametric model for the eyeglass frame element 24, then Figure 5 Method 4' described herein can be used to generate a parametric equivalent model. To this end, in step 14, at least one base entity 38 of the parametric equivalent model of the eyeglass frame element 24 is determined by selecting entities 30 (i.e., designated entities) from the set of entities 30. In subsequent step 16, the base entity 38 is decomposed into the set of segments 40. Subsequently, the designated entity 30 is also decomposed into segments 40 within the set of segments 40. Then, a parametric deformation mapping is selected in such a way that the reconstruction error on the designated entity 30 is minimized. The steps for determining the base entity, segmenting the base entity, and determining the parametric deformation mapping are iterated until a desired quality criterion is met, in the form of the maximum deviation between the surface points 28 of the entities 30 in the set of entities 30 and the surface points 28 of the corresponding entities 30 represented based on the parametric equivalent model.

[0239] Because the specific deformation mapping is determined independently of other segments 40 for each segment 40 of the at least one base entity 38 of the eyeglass frame element 24, discontinuities 78 may exist at segment boundaries 41. These discontinuities can be prevented by a smoothing method applied in post-processing step 46 to the generated mesh 26 of the entity 30, for example, based on the following... Figure 26A , Figure 26B and Figure 26CThe Delta-Mush method is described. Therefore, additional method steps for determining the post-processing method, specifically the smoothing method, of the entity 30 generated based on the parametric equivalent model are advantageous.

[0240] Figure 13 The parametric equivalent model of the eyeglass frame element 24 shows the arrangement of segments 40 in this group of segments 40, in this case, the entire eyeglass frame. The segments 40 are arranged in a hierarchical tree structure 54 according to their spatial relationships. Interconnected nodes 56, 56' indicate the spatial adjacency of the segments 40, meaning that these segments 40 have a common cutting edge or cutting surface. In this case, each segment 40 in the tree structure 54 is positioned and oriented relative to its parent node in coordinate system 32.

[0241] The right subtree 58 of the "nose bridge" node describes the right part of the parametric model of the eyeglasses frame up to the nose bridge, and the left subtree 58 describes the left part up to the nose bridge. The two subtrees 58 of the nose bridge node are symmetrical because the two halves of the parametric model of the eyeglasses frame are also symmetrical.

[0242] If there are multiple eyeglass frame elements 24, these eyeglass frame elements can also be arranged hierarchically in the tree structure 54, for example... Figure 13 As shown.

[0243] Figure 14 A computer-implemented method is described for personalizing the eyeglass frame element 24 by adapting a parametric model of the eyeglass frame element 24 to the head of an eyeglass wearer using the method described above for determining a parametric equivalent model of the eyeglass frame element 24, the parametric equivalent model having at least one parameter. In this case, the representation of the head in coordinate system 32 is determined in a computer unit. Further, the parameter value of the at least one parameter of the parametric equivalent model of the eyeglass frame element 24 is determined such that the entity 30 of the parametric equivalent model of the eyeglass frame element 24 generated based on the at least one parameter value is adapted to the head.

[0244] To this end, the eyeglass frame manufacturer creates a CAD model 22 of the eyeglass frame element 24 with modifiable parameters. To determine the parametric equivalent model of the eyeglass frame element 24 for this model, a set of designated entities 30 are created for various parameter groups in the CAD model. Based on the method described above, the parametric equivalent model of the eyeglass frame element 24 is calculated based on the designated entities 30. This parametric equivalent model of the eyeglass frame element 24 can be stored in a database 42. Then, the corresponding parametric equivalent models of different CAD models 22 of different eyeglass frame elements 24 can be stored in the database 42.

[0245] For example, the database 42, which has a parametric equivalent model of the eyeglass frame element 24, can be used in a system for personalizing and adapting the eyeglass frame element 24 as follows:

[0246] In the first step 48, a representation of the glasses wearer's head is created using a 3D measurement system based on the head model in the coordinate system. A representation of a specific set of parameters is generated for each parametric equivalent model in database 42. This representation can also be stored in database 42 along with the parametric equivalent model to save computation time.

[0247] In a further step 49, the eyeglass wearer can select a frame element 24 from a representation of a parametric equivalent model of various frame elements 24. Based on the parametric equivalent model, the frame element 24 can be adapted to a previously created head model in step 50 using, for example, the algorithms described in EP 3 425 447 A1 or EP 3 425 446 A1, the entire contents of which are incorporated herein by reference and the disclosures of which are included in the specification of this invention.

[0248] To this end, a base entity 38 is selected and decomposed into segments 40 of a parametric equivalent model of the eyeglass frame element 24. This is transformed into the coordinate system 32 of the head model. Finally, the parameters of the parametric deformation mapping of each segment 40 of the base entity 38 are optimized in a way that the eyeglass frame element 24 is adapted to the head model.

[0249] It should be noted that, in principle, in step 50, the parametric equivalent model can also be used to optimally fit the eyeglass frame element 24 to the previously created head model based on user input via the user interface. Parameter values ​​determined for the eyeglass wearer during this process are stored.

[0250] Then, the mesh 26 of the eyeglass frame element 24 is calculated for the selected parametric equivalent model of the eyeglass frame element 24 and the optimized parameter values ​​of the model. This can be indicated in step 52 in the wearing position on the head model of the eyeglass wearer.

[0251] Alternatively, the parameter values ​​of the parametric equivalent model of the eyeglass frame element 24 or the position of the rendered eyeglass frame element 24 can be adapted to the head model.

[0252] The selected eyeglass frame element 24 can then be transferred to the ordering system.

[0253] If the various eyeglass frame components 24 are stored together with their parametric equivalent models in the ordering system, then all that needs to be transmitted for an order is the calculated parameter values ​​of the parametric equivalent models (that is, the index of the selected base entity 38 when the model contains multiple base entities) and the parameter values ​​of the deformation mapping, thereby saving transmission time and even in the case of low bandwidth Internet connection.

[0254] based on Figure 15 A computer-implemented method is described for representing and / or compressing a given entity 30 of a parameterized model of an eyeglass frame element 24, determined in the method described above, in a computer unit. This parameterized equivalent model has at least one parameter. In this case, in a first step, a parameter value is determined for each parameter of the parameterized equivalent model by a group optimization criterion based on a weighted sum, mean, maximum, and quantile of the deviation distribution between the surface of the given entity 30 of the parameterized model and the surface of the entity 30 of the parameterized equivalent model generated based on the at least one parameter value. The determined at least one parameter value is stored in the memory of the computer unit. Figure 15 As shown, for this measurement, the deviation between a given entity 30 and an entity 30 that can be generated based on a parametric equivalent model is minimized by determining optimal parameter values. These parameter values ​​are stored in the memory of the computer unit. In this case, the entity 30 that can be generated based on the parametric equivalent model is obtained by decomposing the parametric model into the group of segments 40 and by parametric deformation mapping. … It is generated by applying to each of the basic segment entities 39 of the n segments.

[0255] Figures 16 to 22 It describes how an algorithm can be used to automatically determine the decomposition of entity 30 of the parametric model or parametric equivalent model of eyeglass frame element 24 into segments 40 in the group of segments 40.

[0256] The algorithm includes the following steps: projecting the surface points 28 of the mesh 26 of entity 30 onto plane 60; determining the signals 72 at the projection points 62 associated with the mirror rings 68 and 70 of entity 30; and determining the inflection points 74 and average values ​​of some signals 76 and 76' of these signals 72. Then, a parameter set of n elements can be used. To describe the decomposition.

[0257] In this example of the front of the frame, the group of segments 40 consists of twelve segments 40. To make the segmentation method applicable to various entities 30 of the same parametric model, the entities 30 are aligned in coordinate system 32, such as based on... Figure 7 As described.

[0258] The surface points 28 of the mesh 26 of the entity 30 to be decomposed are projected onto the plane 60 along the spatial axis, as shown below. Figure 16 As shown. Projection points 62, which are in the form of projection points, can be classified along an axis, in this case, the horizontal axis.

[0259] Select two groups from projection point 62: the first group 64 contains... Figure 17 The projection point 62 of surface point 28 on the upper lens ring 68 of the solid 30 of the CAD model of the front of the eyeglass frame shown. The second group 66 contains Figure 17 The projection point 62 of the lower lens ring 70 surface point 28 of the solid 30 of the front part of the frame CAD model.

[0260] For example, the horizontal coordinate can be sensed at regular intervals, such as 1 mm.

[0261] To obtain the first set of 64 projection points 62, a set of projection points 62 with similar horizontal coordinate values ​​can be determined for each sensed value on the horizontal axis, and the projection point 62 with the maximum value on the vertical axis can be selected from them.

[0262] To obtain the second set of 66 projection points 62, a set of projection points 62 with similar horizontal coordinate values ​​can be determined for each sensed value on the horizontal axis, and the projection point 62 with the minimum value on the vertical axis can be selected from them.

[0263] Then, based on the first group 64 and the second group 66 projection points 62 of the parametric model of the eyeglass frame element 24, the entity 30 can be automatically decomposed into segments 40 in the group of segments 40 by an algorithm.

[0264] The upper mirror circle 68, represented by the first set of 64 projection points 62 as contours in plane 60, and the lower mirror circle 70, represented by the second set of 66 projection points 62 as contours in plane 60, can be considered as signals 72. For the decomposition of these signals, signal processing algorithms can be used, such as the algorithm for detecting inflection points 74 described in the article "Using penalized contrasts for the change-point problem," Marc Lavielle, Signal Processing, 2005, Vol. 85, pp. 1801-1810, the entire contents of which are incorporated herein by reference and the disclosures of which are included in the specification of this invention.

[0265] If the signal is 72 pixels Figure 18 If the signal shown is also usable, then the inflection point 74 of signal 72 can be automatically determined based on this algorithm. Let... For sampling points Values ​​used at (horizontal axis) Continuous signal 72 (vertical axis). Based on the following optimization problem, it is possible to minimize the number of signals including... The first part of the signal 76 and contains The objective function of the sum of variances of the second part of the signal 76' To calculate the contents Inflection point 74 in segment 72 of signal 72:

[0266]

[0267] Problem (1) can be modified to be optimized in such a way that any desired number of inflection points 74 can be detected in signal 72.

[0268] Figure 19A , Figure 19B and Figure 19C The calculation of inflection point 74 in signal 72 from projection points 62 of the first group 64 and the second group 66 is shown. In each case, the vertical line shows the coordinates of the inflection point 74 detected in signal 72. The horizontal line shows the average value of partial signals 76 and 76'. .

[0269] exist Figure 19A In the middle, based on the lower mirror ring 70 described by the second set of 66 projection points 62, an optimization problem (1) is solved for the four inflection points 74, and thus the sum of the variances of the five partial signals 76, 76' is minimized. In the xz plane 60, the horizontal axis shows the index i of the projection points 62 from the second set of 66 projection points 62, which are associated with the surface point 28 of the lower mirror ring 70. The vertical axis shows the z-coordinate of the projection point 62.

[0270] Figure 19B yes Figure 19A The signal 72 segment, especially the second group of 66 projection points 62 in the interval In this portion, these projection points are located on the lower lens ring 70 of the nose bridge via projection surface point 28. In the xz plane 60, the horizontal axis shows the index i of the projection points 62 from the second group of 66 projection points 62, which are associated with surface point 28 of the lower lens ring 70. The vertical axis shows the z-coordinate of the projection point 62. For this signal segment, two inflection points 74 are detected again in subsequent steps.

[0271] Figure 19C The determination of the four inflection points 74 of the first group 64 projection points 62 of the upper lens ring 68 is shown.

[0272] The decomposition of solid 30 of the parametric model of the front of the eyeglass frame can be described, for example, by the following set of parameters.

[0273]

[0274] Among them, there are 16 parameter values:

[0275] • x 1: The minimum x-coordinate of all projected points 62

[0276] • x 9: The maximum x-coordinate of all projected points 62

[0277] • z 1: The minimum ordinate of all projected points 62

[0278] • z 4: The maximum ordinate of all projected points 62

[0279] •

[0280] • x 3: [ x 1, x The x-coordinate of the minimum y-coordinate in 5]

[0281] • x 32 :[ x 1, x The x-coordinate of the largest y-coordinate in 5]

[0282] • x 7: [ x 5, x The x-coordinate of the minimum ordinate in 9]

[0283] • x 71 :[ x 5, x The x-coordinate of the largest y-coordinate in 9]

[0284] • z 2: M 1

[0285] • z 21 : M 5

[0286] • x 4: C 5

[0287] • x 6: C 6

[0288] • z 3: M 7

[0289] •x 2: C 7

[0290] • x 8: C 10 .

[0291] Figure 20A The figure shows the decomposition of the entity 30 of the parametric equivalent model of the front of the frame into twelve segments 40 determined based on the algorithm described above for detecting the inflection point 74. Figure 20B The parametric equivalent model of the front of the frame is shown to be decomposed into twelve segments 40 calculated based on the same algorithm. In this case, all surface points 28 located within the numerically marked areas are part of the same segment 40 with segment boundaries 41. Figure 20A and Figure 20B The segments 40 of the two entities 30, which are marked with the same number, correspond to each other.

[0292] Figure 21 The diagram shows the decomposition of the solid 30 of the CAD model of the temple into two segments 40.

[0293] Since the same decomposition algorithm is applied to all entities 30 of the parametric model of the at least one eyeglass frame element 24 or the parametric equivalent model of the at least one eyeglass frame element 24, each segment 40 of one entity 30 can be directly assigned to a corresponding segment 40 in another entity 30. Based on these correspondences, a parametric deformation mapping for mapping the base segment entity 39 to another corresponding segment entity 43 can be determined.

[0294] To improve the accuracy of parametric deformation mapping, the parameter values ​​in (2) can be changed. Z To optimize the segmentation of entity 30, such as Figure 22 As shown. This can improve the ability to map different segment entities 43 on the same segment 40 to each other.

[0295] As an alternative to detecting the inflection point 74 in signal 72 from mirror rings 68 and 70, in order to determine the parameter set in (2) Z To decompose entity 30 into segments 40 within the group of segments 40, mesh segmentation, multivariate adaptation, skeletonization, and / or machine learning methods can be used.

[0296] In order to represent the entity 30 of the parametric model of the eyeglass frame element 24 based on the parametric equivalent model of the eyeglass frame element 24, after decomposing the entity 30 into segments 40 in the group of segments 40, the parameter values ​​of the associated parametric deformation mapping of each of these segments 40 must be determined.

[0297] Therefore, algorithms for aligning 3D objects that minimize the distance between point clouds can be used, such as those described in the article " S. Rusinkiewicz and M. Levoy, Efficient variants of the ICP algorithm, Proceedings of the Third International Conference on 3-D Digital Imaging and Modeling [Proceedings of the 3rd International Conference on 3D Digital Imaging and Modeling] Pages 145-182, 2001 The Iterative Closest Point (ICP) algorithm described in the document is cited in its entirety and the disclosures of that document are included in the specification of this invention.

[0298] It can be assumed that when the parameter values ​​of the parametric deformation map change, the triangulation of surface points 28 in the form of a triangular mesh, specifically the topology and links of the triangular structure, remain unchanged. The algorithm used to determine the parameter values ​​of the parametric deformation map (e.g., the ICP algorithm) can then be directly applied to surface points 28 of mesh 26. This saves computation time.

[0299] Figure 23A The deformation of the base segment entity 39 of the parametric model of the temple of the eyeglass based on the ICP algorithm is shown, such that the distance between the surface point 28 of the mesh 26 of the base segment entity 39 and the surface point 28 of the mesh 26 of the corresponding segment 40 in another entity 30 of the parametric model of the temple is as small as possible.

[0300] Figure 23A The surface points 28 of the mesh 26 of the base segment entity 39 and the other base segment entity 43 in coordinate system 32 before the application of the ICP algorithm are shown. Figure 23B The two segments 40 are shown after the algorithm has iterated 18 times.

[0301] Figure 23C The root mean square error curves of the shortest distance between surface points 28 of the base segment entity 39 and the other segment entity 43 are shown.

[0302] For some eyeglass frame elements 24, such as the temples, the parametric deformation mapping of segment 40 can be selected particularly easily, for example, simply as a combination of rotation matrices and translation vectors. The parameter values ​​can then be determined based on the ICP algorithm.

[0303] In this case, the following form of mapping can be selected.

[0304]

[0305] As a parametric deformation mapping, where This represents a special orthogonal group of all rotations about the origin in three-dimensional Euclidean space.

[0306] In this case, the following optimization problem is solved iteratively, which minimizes the surface points of the mesh 26 of the basic segment entity 39.p i The closest to the mesh 26 of the corresponding segment 40 of the other entity 30 p i Surface points q i The weight of the distance is w i Weighted sum:

[0307]

[0308] You can choose the weight as w i = 1. Alternatively, other weights are also applicable. For example, a point can be determined based on the angle between the surface normals present at that point. p i and weight :

[0309]

[0310] The surface normal of a point can be estimated based on its nearest neighbors in the point cloud. For example, this type of weighting is described in the aforementioned article on the ICP algorithm.

[0311] Alternatively, other ICP variants also apply, such as those mentioned in the article " Paul J. Besl and Neil D. McKay, A Method for Registration of 3-D Shapes [Registration methods for 3D shapes] IEEE Transactions on Pattern Analysis and Machine Intelligence [IEEE Transactions on Pattern Analysis and Machine Intelligence] Volume 14, 2nd edition, 1992 The entire contents of the article described herein are quoted hereafter and the disclosures of the article are included in the specification of this invention.

[0312] The point-to-plane ICP algorithm is advantageous for the speed of this method, for example, in the article " Kok-Lim Low, Linear Least-Squares Optimization for Point-to-Plane ICP Surface Registration [Linear Least Squares Optimization for Point-to-Plane ICP Surface Registration] Department of Computer Science, University of North Carolina at Chapel Hill February 2004 The entire contents of the article described herein are cited and the disclosures of that article are included in the specification of this invention. In this case, what is minimized is not the distance between surface points of entities, but the distance between a surface point of one entity and the tangent plane at the nearest surface point of another entity.

[0313] Figure 24AExamples of parameter values ​​for the parametric deformation mapping of different base segment entities 39 of the parametric model used to determine the front of the eyeglass frame are shown. In this case, the deviation of the surface point 28 of the mesh 26 of the base segment entity 39 from the nearest surface point 28 of the mesh 26 of the segment 40 of the other entity 30 is minimized based on the optimization problem (3).

[0314] Figure 24A The surface points 28 of the base segment entity 39 and the additional entity 30 before the application of the ICP algorithm are shown. Figure 24B The minimum deviations of segments 40 listed in 1 to 6 are shown. Figure 24C The minimum deviation of all segments 40 is shown after the parameter values ​​of the deformation mapping are determined.

[0315] As an alternative to the ICP algorithm, other deformation methods, specifically mesh editing methods, can be used to determine the parameter values ​​for the parameterized deformation map. In this case, the surface is deformed based on control points by solving a sparse matrix problem. Examples of mesh editing methods include, for instance, those described in the article " Laplacian Surface Editing [Laplace Surface Editing] , O Sorkine, D. Cohen-Or, Eurographics Symposium on Geometry Processing [European Symposium on Geometric Processing in Computer Graphics], 2004 Laplace surface editing as described in the article “Mesh Editing with Poisson-Based Gradient Field Manipulation”, Yizhou Yu et al., ACM SIGGRAPH 2004, or Poisson surface editing as described in the article “Mesh Editing with Poisson-Based Gradient Field Manipulation”, Yizhou Yu et al., ACM SIGGRAPH 2004.

[0316] Advantageously, the symmetry assumptions of individual segments 40 (e.g., left temple and right temple) of the parametric model of the at least one eyeglass frame element 24 are integrated into the parametric equivalent model of the at least one eyeglass frame element 24.

[0317] With relatively small changes to the parametric model of the eyeglass frame element 24, the complexity of the parametric equivalent model can be reduced by using a larger set of base entities 38 instead of parametric deformation mapping.

[0318] The parametric equivalent model determined for the parametric model of the front of the eyeglass frame can include the following elements with parameters:

[0319] - Multiple segments 40 of the parametric model of the front of the frame;

[0320] - The grid 26 of the at least one basic segment entity 39 at the front of the frame;

[0321] - (2) parameter groupZ This parameter group contains 16 parameters that describe the boundaries of the twelve segments 40;

[0322] - It has 12 rotation matrices and 12 translation vectors with parameters to be determined, which describe the parametric deformation mapping of each segment 40 of the basic segment entity 39;

[0323] - Parameters for post-processing step 46.

[0324] In this case, the parameter set in (2) describing the parametric model decomposition Z It is optional because it must always be recalculated based on the decomposition algorithm and therefore does not need to be stored like the parameters of the parameterized equivalent model. This saves transfer time and memory space. However, additional storage saves computation time.

[0325] For a specific entity 30 of the parametric equivalent model of the front of the eyeglass frame, storing the following parameter values ​​is sufficient:

[0326] - The index of the selected basic segment entity 39 for each segment 40 in the parametric equivalent model when multiple basic segment entities 39 can be used for a segment 40;

[0327] - Parameter values ​​for parametric deformation mapping.

[0328] These parameter values ​​can be transmitted to a video centering device. At the video centering device, a specific entity 30 can be recovered based solely on the corresponding index of the base segment entity 39 and the parameter values ​​of the parametric deformation mapping, and based on the parametric equivalent model of the at least one eyeglass frame element 24 (which is stored at the video centering device). Therefore, it is not necessary to transmit the entire mesh 26 of the parametric equivalent model of the specific entity 30 or the at least one eyeglass frame element 24 to the video centering device and store it there. Thus, the use of the parametric equivalent model saves memory space and transmission time.

[0329] Based on the parametric equivalent model, the mesh 26 of the eyeglass frame element 24 can be generated by selecting parameter values, such as... Figure 25 As shown.

[0330] To avoid discontinuities 78 at segment boundaries 41 that may occur due to the independent calculation of parameter values ​​for parametric deformation mapping on each segment 40, a smoothing method, such as the Delta Mush method described in the article "Delta Mush: Smoothing Deformations while Preserving Detail," Joe Mancewicz, Matt L. Derksen, Hans Rijpkema, Cyrus A. Wilson, Proceedings of the 4th Symposium on Digital Production, 2014, the entire contents of which are incorporated herein by reference and the disclosures of which are included in this specification.

[0331] The Delta Mush method is superior to other smoothing methods in that, due to the smoothing, the mesh 26 exhibits only minor changes, and therefore, calculations requiring particularly high accuracy (e.g., virtual centering) can be performed even using a parametric equivalent model of at least one eyeglass frame element 24.

[0332] Figure 26A , Figure 26B and Figure 26C The entity 30, which uses a parameterized equivalent model of the connection points, explains how the DeltaMush method is applied to the segment boundary 41. Figure 26A The entity 30 is shown as a parameterized equivalent model of the connection point with discontinuity 78 at the segment boundary 41. Figure 26B The segment 40, after being smoothed using the Delta Mush method, is shown as having no discontinuities 78. For comparison purposes, Figure 26C The original entity 30 of the parameterized model of the connection point is shown.

[0333] The computer program product according to the present invention comprises a computer program having program code for performing the above-described method steps when the computer program is loaded into a computer unit and / or executed on the computer unit.

[0334] An apparatus for personalizing and adapting a parametric model of an eyeglass frame element to the head of an eyeglass wearer includes a computer unit loaded with a computer-implemented method for adapting the parametric model of the eyeglass frame element to a head representation in a coordinate system.

[0335] An apparatus for representing and / or compressing a parametric model of a given entity of an eyeglass frame element includes a computer unit having a memory in which a computer-implemented method for representing and / or compressing the given entity in the memory of the computer unit is loaded.

[0336] The system according to the invention having a device for performing the following operations uses at least one of the parameter values ​​determined by the parameterized equivalent model: producing personalized eyeglass frame elements in the method for personalizing eyeglass frame elements described above, or grinding eyeglass lenses into eyeglass frame elements as personalized in the method for personalizing eyeglass frame elements described above.

[0337] In summary, particular attention should be paid to the following: The present invention relates to a method 10, 10' for determining a parametric equivalent model of an eyeglass frame element 24 for a parametric model in order to adapt the parametric equivalent model to the head of an eyeglass wearer. In this case, at least one base entity 38 is provided by creating at least one entity 30 of the parametric model of the at least one eyeglass frame element 24 based on a set of specific parameter values ​​to realize the parametric model of the at least one eyeglass frame element 24. At least one parametric deformation mapping is determined for the at least one base entity 38, which maps the at least one base entity 38 onto the entity 30 of the parametric model, and the parametric equivalent model is determined at least based on the at least one base entity 38 and the at least one parametric deformation mapping. Alternatively, a set of segments 40 may be determined for the parametric model of the eyeglass frame element 24. At least one base segment entity 39 and at least one parametric deformation mapping are determined for each segment 40. The at least one parametric deformation mapping then maps at least one base segment entity 39 onto another segment entity 43 of the parametric model, determining the parametric equivalent model at least based on the group of segments 40 and based on the at least one base segment entity 39 and the at least one parametric deformation mapping for each segment 40 in the group of segments 40.

[0338] Preferred features of the present invention:

[0339] 1. A computer-implemented method (10) for determining a parametric equivalent model of a parametric model of an eyeglass frame element (24), the parametric equivalent model having at least one parameter, wherein,

[0340] Multiple entities of the parameterized model are specified in the form of implementing the parameterized model through specific parameter values ​​(30).

[0341] Based on these designated entities (30), at least one basic entity (38) and

[0342] At least one parametric deformation mapping of the at least one base entity (38), the at least one parametric deformation mapping mapping the at least one base entity (38) onto the entity (30) of the parametric model, and determining the parametric equivalent model based at least on the at least one base entity (38) and the at least one parametric deformation mapping.

[0343] 2. A computer-implemented method (10) for determining a parametric equivalent model of a parametric model of an eyeglass frame element (24), the parametric equivalent model having at least one parameter, wherein,

[0344] Multiple entities of the parameterized model are specified in the form of implementing the parameterized model through specific parameter values ​​(30).

[0345] A set of segments (40) are determined for the parametric model of the eyeglass frame element (24).

[0346] These specified entities (30) are decomposed into segments (40) in the group of segments (40).

[0347] For each segment (40) in the group of segments (40), a segment entity (43) is generated by means of the entity (30) of that segment (40) selected from the specified entity (30) being decomposed.

[0348] Based on these segment entities (43), at least one basic segment entity (39) is determined and

[0349] At least one parametric deformation mapping of the at least one basic segment entity (39),

[0350] The at least one parametric deformation mapping maps the at least one basic segment entity (39) onto the segment entity (43) of the parametric model.

[0351] And the parametric equivalent model is determined at least based on the group of segments (40) and the at least one basic segment entity (39) and the at least one parametric deformation mapping of each segment (40) in the group of segments (40).

[0352] 3. A computer-implemented method (10) for determining a parametric equivalent model of a parametric model of an eyeglass frame element (24), the parametric equivalent model having at least one parameter, wherein,

[0353] Multiple entities of the parameterized model are specified in the form of implementing the parameterized model through specific parameter values ​​(30).

[0354] A set of segments (40) are determined for the parametric model of the eyeglass frame element (24).

[0355] These specified entities (30) are decomposed into segments (40) in the group of segments (40).

[0356] For each segment (40) in the group of segments (40), a segment entity (43) is generated by means of the entity (30) of that segment (40) selected from the specified entity (30) being decomposed.

[0357] For each segment (40) in the computer-implemented method according to Clause 1, a parameterized equivalent model having at least one parameter is determined as the segment equivalent model, and the segment entity (43) associated with each segment is used as the designated entity in this context.

[0358] And the parameterized equivalent model is determined at least based on the group of segments (40) and the parameterized equivalent model of the segment with at least one parameter.

[0359] 4. The method according to clause 2 or 3, characterized in that the segments (40) in the group of segments (40) are marked as static, movable or deformable.

[0360] 5. The method according to Clause 4, characterized in that the parametric deformation maps are linear maps of segments (40) marked as static, and / or the parametric deformation maps of segments (40) marked as movable are affine maps, and / or the parametric deformation maps of segments (40) marked as deformable are approximated based on Bézier curves, splines or NURBS.

[0361] 6. The method according to any one of clauses 2 to 5, characterized in that, during the decomposition of the entity (30) of the parameterized model of the eyeglass frame element (24) into segments (40) of the group of segments (40), a method for identifying inflection points (74) in the signal (72) and / or a grid segmentation method and / or a multi-adaptation method and / or a skeletonization method and / or a machine learning method is applied;

[0362] and / or

[0363] The segments (40) in this group of segments (40) are arranged in layers in the tree structure (54) in such a way that the nodes (56, 56') connected in the tree structure (54) are associated with the segments (40) in the parametric model that have a common cutting edge or cutting surface.

[0364] and / or

[0365] Each segment (40) in the tree structure (54) is positioned and oriented relative to its parent segment in the coordinate system (32).

[0366] and / or

[0367] The entity (30) of the parameterized equivalent model is post-processed based on an algorithm for avoiding discontinuities (78) at the segment boundary (41) in the form of a parameterized equivalent model with specific parameter values.

[0368] 7. The method according to any one of Clauses 1 to 6, characterized in that additional features are determined for the parametric equivalent model of the eyeglass frame element (24), the additional features being derived from a group including ear support points, nose support points, support curves at the ends of temples, 3D lens planes, 3D boxes, and nose pads;

[0369] and / or

[0370] This parametric deformation mapping originates from a group including affine mappings, polynomials, polynomial surfaces, Bézier curves, splines, or NURBS.

[0371] and / or

[0372] The method for determining this parameterized equivalent model is iterative.

[0373] 8. The method according to any one of clauses 1 to 7, characterized in that, in order to determine the parametric equivalent model, the criteria are optimized based on a weighted sum, mean, maximum, and quantile set of the deviation distribution between the surface of the specified entity (30) including the parametric model and the surface of all those entities (30) that can be generated based on specific parameter values ​​of the parametric equivalent model of the at least one eyeglass frame element (24).

[0374] and / or

[0375] The specified entity (30) of the parameterized model is post-processed at least in part by algorithms used to correct errors and / or to improve the visual impression of the eyeglass wearer and / or for smoothing.

[0376] 9. To provide a parametric equivalent model determined in any one of clauses 1 to 8 in a data format different from the data format of the parametric model.

[0377] 10. A computer-implemented method for personalizing an eyeglass frame element (24) by adapting a parametric model of the eyeglass frame element (24) to the head of an eyeglass wearer using a parametric equivalent model, the parametric equivalent model having at least one parameter and determined in the method according to any one of clauses 1 to 8 or provided based on clause 9,

[0378] Its characteristics are,

[0379] Determine the representation of the head in the coordinate system (32) of the computer unit; and

[0380] Determine the parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24) such that the entity (30) of the parametric equivalent model of the eyeglass frame element (24) generated based on the at least one parameter value is adapted to the head.

[0381] 11. A computer-implemented method for representing and / or compressing a given entity (30) of a parametric model of an eyeglass frame element (24) in a computer unit, the parametric equivalent model having at least one parameter and determined in the method according to any one of Clauses 1 to 8 or provided based on the method according to Clause 9,

[0382] Its characteristics are,

[0383] The corresponding parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24) is determined by optimizing a criterion based on a weighted sum, mean, maximum, and quantile of the distribution of deviations between the surface of a given entity (30) of the parametric model and the surface of the entity (30) of the parametric equivalent model generated based on the at least one parameter value; and

[0384] At least one determined parameter value is stored in the memory of the computer unit.

[0385] 12. A computer program having program code for performing all method steps as described in any one of clauses 1 to 11 when the computer program is loaded onto a computer unit and / or executed on a computer unit.

[0386] 13. An apparatus for personalizing and adapting a parametric model of an eyeglass frame element (24) to the head of an eyeglass wearer, the apparatus comprising a computer unit including a computer-implemented method, as described in Clause 10, for adapting the parametric model of the eyeglass frame element (24) to a head representation in a coordinate system (32) within the computer unit.

[0387] 14. An apparatus for representing and / or compressing a parametric model of a given entity (30) of an eyeglass frame element (24), the apparatus comprising a computer unit having a memory, the computer unit including a computer-implemented method for representing and / or compressing the given entity in the memory of the computer unit as described in Clause 11.

[0388] 15. An apparatus having at least one determined parameter value for using the parameterized equivalent model to produce a personalized eyeglass frame element (24) or to grind eyeglass lenses into a personalized eyeglass frame element (24) as described in accordance with Clause 10.

[0389] List of reference numerals

[0390] 2. Method and Steps: Specify the parametric model of the eyeglass frame components

[0391] 4, 4', 4" Method and Steps: Determine the parametric equivalent model of the eyeglass frame components.

[0392] 6. Methods and Procedures: Provide biometric data related to the head of the eyeglass wearer.

[0393] 8. Method Steps: Determine at least one parameter value for the parameterized equivalent model by optimizing the function used to adapt the parameterized equivalent model to the head of the eyeglass wearer.

[0394] 10, 10', 10" method

[0395] 12. Method Steps: Specify the solid model of the parametric model of the eyeglass frame element.

[0396] 13. Method and Steps: Decompose the parametric model of the eyeglass frame components into a set of segments.

[0397] 14. Method and Steps: Identify at least one basic entity

[0398] 15. Method Steps: Decompose the specified entity into segments within the group of segments.

[0399] 16. Method Steps: Determine at least one parameterized deformation mapping

[0400] 17. Method Steps: Select segment entities from the specified entities to be decomposed.

[0401] 18. Iterate through the steps of the method used to optimize the parameterized equivalent model.

[0402] 20. Method and Steps: Determine at least one basic segment entity for each segment.

[0403] 21. Method Steps: Determine at least one parametric deformation mapping for each basic segment entity.

[0404] 22 CAD models

[0405] 24 Eyeglass Frame Components

[0406] 26 grids

[0407] 28 Surface Points

[0408] 30 entities

[0409] 31 Biometric Data

[0410] 32 coordinate system

[0411] 34. Plane of Symmetry

[0412] 36. Center of mass

[0413] 38 Basic Entities

[0414] 39. Basic Section Entity

[0415] 40 sections

[0416] 41 Section Boundaries

[0417] 42 Databases

[0418] 43 Section Entities

[0419] 44 Preprocessing Steps

[0420] 46 Post-processing steps

[0421] 48. Method and Steps: Generating the Head Model

[0422] 49. Method Steps: Select the base entity

[0423] 50. Method and Steps: Adapt the parametric equivalent model to the head of the glasses wearer.

[0424] 52. Method and Steps: Virtual Wearing and Rendering of Parametric Model Entities

[0425] 54 Tree Structure

[0426] Nodes 56 and 56'

[0427] 58 subtrees

[0428] 60 plane

[0429] 62 projection points

[0430] 64 First group of projection points

[0431] 66 Second group of projection points

[0432] 68 upper lens circle

[0433] 70mm lower lens circle

[0434] 72 signals

[0435] 74 Turning Point

[0436] 76, 76' partial signals

[0437] 78 Discontinuity

[0438] Coordinates of the detected inflection point

[0439] Average value of the signal between inflection points

[0440] f、f i Parametric deformation mapping

[0441] α、α i Parameters of parametric deformation mapping

[0442] b Basic Entity

[0443] b i Basic segment entity.

Claims

1. A computer-implemented method (10) for personalizing an eyeglass frame element (24) by adapting a parametric model of the eyeglass frame element (24) to the head of an eyeglass wearer. Its characteristics are, The parametric equivalent model of the parametric model of the eyeglass frame element (24) is determined by means of the following, which has at least one parameter: Multiple entities (30) of the parameterized model are specified in the form of implementing the parameterized model through specific parameter values, wherein the entities of the parameterized model are specific examples of the implementation of the parameterized model for the parameter values ​​selected for the parameters of the parameterized model. Based on these designated entities (30), at least one basic entity (38) and At least one parameterized deformation mapping of the at least one base entity (38), wherein the base entity is a specific entity of the parameterized model that is selected or computed and used to define the parameterized deformation mapping, the at least one parameterized deformation mapping maps the at least one base entity (38) to the entity (30) of the parameterized model, and the parameterized equivalent model is determined at least based on the at least one base entity (38) and the at least one parameterized deformation mapping; Provide biometric data related to the head of the person wearing the glasses; as well as The function is optimized to determine at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24), which takes into account at least one surface point of the determined base entity (38) of the parametric equivalent model of the eyeglass frame element (24) and the biometric data (31) related to the head of the eyeglass wearer.

2. A computer-implemented method (10') for personalizing the eyeglass frame element (24) by adapting a parametric model of the eyeglass frame element (24) to the head of an eyeglass wearer. Its characteristics are, The parametric equivalent model of the parametric model of the eyeglass frame element (24) is determined by means of the following, which has at least one parameter: Multiple entities (30) of the parameterized model are specified in the form of implementing the parameterized model through specific parameter values, wherein the entities of the parameterized model are specific examples of the implementation of the parameterized model for the parameter values ​​selected for the parameters of the parameterized model. A set of segments (40) is determined for the parametric model of the eyeglass frame element (24), wherein these segments are a subset of the eyeglass frame element. These specified entities (30) are decomposed into segments (40) in the group of segments (40). For each segment (40) in the group of segments (40), a segment entity (43) is generated by means of the entity (30) of that segment (40) selected from the specified entity (30) to be decomposed, wherein the segment entity represents the entity of the segment of the parameterized model. Based on these segment entities (43), at least one basic segment entity (39) is determined and At least one parametric deformation mapping of the at least one basic segment entity (39), The at least one parametric deformation mapping maps the at least one basic segment entity (39) onto the segment entity (43) of the parametric model. And the parametric equivalent model is determined at least based on the group of segments (40) and based on the at least one basic segment entity (39) of each segment (40) in the group of segments (40) and the at least one parametric deformation mapping; Provide biometric data related to the head of the person wearing the glasses; as well as The function is optimized to determine at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24), which takes into account at least one surface point of at least one determined base segment entity (39) of the parametric equivalent model of the eyeglass frame element (24) and the biometric data (31) related to the head of the eyeglass wearer.

3. A computer-implemented method (10") for personalizing the eyeglass frame element (24) by adapting a parametric model of the eyeglass frame element (24) to the head of an eyeglass wearer. Its characteristics are, The parametric equivalent model of the parametric model of the eyeglass frame element (24) is determined by means of the following, which has at least one parameter: For the parametric model of the eyeglass frame element (24), a set of segments (40) are determined, and for each segment (40), a parametric segment model from the parametric model of the eyeglass frame element (24) is determined. For each parameterized segment model in the computer-implemented method as described in claim 1, a parameterized equivalent model having at least one parameter is determined as the segment equivalent model. And the parameterized equivalent model is determined at least based on the group of segments (40) and the parameterized equivalent model of the segment with at least one parameter; Provide biometric data related to the head of the person wearing the glasses; as well as The function is optimized to determine at least one parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24), the function taking into account at least one surface point of the determined base entity (38) of at least one segment of the parametric equivalent model of the eyeglass frame element (24) and the biometric data (31) related to the head of the eyeglass wearer.

4. The method as described in claim 2 or 3, characterized in that, The segments (40) in this group are marked as static, movable or deformable.

5. The method as described in claim 4, characterized in that, These parametric deformation maps are linear maps of segments (40) labeled as static, and / or these parametric deformation maps of segments (40) labeled as movable are affine maps, and / or these parametric deformation maps of segments (40) labeled as deformable are approximated based on Bézier curves, splines or NURBS.

6. The method as described in claim 2 or 3, characterized in that, During the decomposition of the entity (30) of the parameterized model of the eyeglass frame element (24) into segments (40) in the group of segments (40), methods for identifying inflection points (74) in the signal (72) and / or mesh segmentation methods and / or multi-adaptation methods and / or skeletonization methods and / or machine learning methods are applied; and / or The segments (40) in this group of segments (40) are arranged in layers in the tree structure (54) in such a way that the nodes (56, 56') connected in the tree structure (54) are associated with the segments (40) in the parametric model that have a common cutting edge or cutting surface. and / or Each segment (40) in the tree structure (54) is positioned and oriented relative to its parent segment in the coordinate system (32). and / or The algorithm based on avoiding discontinuities (78) at the segment boundary (41) performs post-processing on the entity (30) of the parameterized equivalent model implementation form using specific parameter values.

7. The method according to any one of claims 1 to 3, characterized in that, Additional features are determined for the parametric equivalent model of the eyeglass frame element (24), which are derived from a group including ear support points, nose support points, support curves at the ends of temples, 3D lens planes, 3D boxes, and nose pads. and / or This parametric deformation mapping originates from a group including affine mappings, polynomials, polynomial surfaces, Bézier curves, splines, or NURBS. and / or The method for determining this parameterized equivalent model is iterative.

8. The method according to any one of claims 1 to 3, characterized in that, To determine the parametric equivalent model, the criteria are optimized based on a weighted sum, mean, maximum, and quantile of the distribution of deviations between the surfaces of the specified entities (30) including the parametric model and the parametric equivalent model of the at least one eyeglass frame element (24) that can be generated based on specific parameter values. and / or The specified entity (30) of the parameterized model is post-processed at least in part by algorithms used to correct errors and / or to improve the visual impression of the eyeglass wearer and / or for smoothing.

9. The method according to any one of claims 1 to 3, characterized in that, The biometric data (31) associated with the head of the eyeglass wearer consists of a representation of the eyeglass wearer’s head, specifically at least one surface point of a grid.

10. The method as described in claim 9, characterized in that, The function to be optimized minimizes the distance between point clouds, wherein the first point cloud contains at least one surface point of the base entity (38) of the parameterized equivalent model of the eyeglass frame element (24) and the second point cloud contains at least one surface point of the representation of the head of the eyeglass wearer.

11. The method according to any one of claims 1 to 3, characterized in that, The parameterized equivalent model is provided in a data format that is different from the data format of the parameterized model.

12. A computer-implemented method for representing and / or compressing a given entity (30) of a parametric model of an eyeglass frame element (24) in a computer unit, the parametric equivalent model having at least one parameter and determined in the method of any one of claims 1 to 10 or provided based on the method of claim 11. Its characteristics are, The corresponding parameter value of at least one parameter of the parametric equivalent model of the eyeglass frame element (24) is determined by optimizing a criterion based on a weighted sum, mean, maximum, and quantile set of the deviation distribution between the surface of a given entity (30) of the parametric model and the surface of the entity (30) of the parametric equivalent model generated based on the at least one parameter value; and At least one determined parameter value is stored in the memory of the computer unit.

13. A computer program having program code for implementing all the method steps specified in any one of claims 1 to 12 when the computer program is loaded onto a computer unit and / or executed on a computer unit.

14. An apparatus for personalizing and adapting a parametric model of an eyeglass frame element (24) to the head of an eyeglass wearer, the apparatus comprising a computer unit including a computer-implemented method as claimed in any one of claims 1 to 10 for adapting the parametric model of the eyeglass frame element (24) to the head of the eyeglass wearer in the computer unit.

15. An apparatus for representing and / or compressing a parametric model of a given entity (30) of an eyeglass frame element (24), the apparatus comprising a computer unit having a memory, the computer unit including the computer-implemented method of claim 12 for representing and / or compressing the given entity in the memory of the computer unit.

16. A system having means for generating an eyeglass frame element (24) personalized using at least one determined parameter value of a parametric equivalent model in the method of any one of claims 1 to 10, or for grinding eyeglass lenses to an eyeglass frame element (24) personalized using at least one determined parameter value of a parametric equivalent model as described in any one of claims 1 to 10.

Images