Method, device, and computer program for virtually adjusting a spectacle frame

Abstract

A virtual try-on process for spectacles includes an approximate positioning and a fine positioning of a spectacle frame on a head of a user. Provided for this purpose are 3D models of the head and the spectacle frame, as well as head metadata based on the model of the head and frame metadata based on the model of the frame. The head metadata contains placement information, in particular a placement point, which can be used for the approximate positioning of the spectacle frame on the head, and/or a placement region which describes a region of the earpiece part of the frame for placement on the ears of the head. A rapid and relatively simple computational positioning of the spectacle frame on the head and a more accurate positioning using a subsequent precise adjustment can be achieved with the aid of the metadata.

Claims

1. A computer-implemented method for virtual fitting of a pair of spectacles, the method comprising: defining a horizontal direction corresponding to a connecting line between pupil centers of eyes of a head in a main fixation direction when the head is held erect as an x-direction, a vertical direction when the head is held erect as a y-direction, and a direction perpendicular to the x-direction and to the y-direction as a z-direction; performing an approximate virtual positioning of a spectacle frame on the head on a basis of head metadata for a 3D model of the head and frame metadata for a 3D model of the spectacle frame; subsequently performing a fine virtual positioning of the spectacle frame on the basis of the 3D model of the head and the 3D model of the spectacle frame; displacing, during the fine virtual positioning, the spectacle frame by a first distance perpendicular to the z-direction in a first displacement; and displacing, during the fine virtual positioning, the spectacle frame by a second distance along the z-direction in a second displacement, wherein the second distance is no more than 10% of the first distance, wherein the approximate positioning determines the position of the spectacle frame in the z-direction, and wherein the position of the spectacle frame in the z-direction remains fixed during the fine positioning.

2. The method as claimed in claim 1, wherein the approximate positioning determines the position of the spectacle frame in the z-direction.

3. The method as claimed in claim 1, wherein providing the frame metadata comprises at least one of identifying prominent points in the 3D model of the spectacle frame or classifying the prominent points in the 3D model of the spectacle frame.

4. The method as claimed in claim 3, further comprising: applying a coordinate transformation on the basis of the identified prominent points to at least one of the group of: the 3D model of the spectacle frame, the identified prominent points, or the frame metadata.

5. The method as claimed in claim 1, wherein at least one of the frame metadata includes first resting information items, which define one or more locations on the spectacle frame where the spectacle frame rests on the head, or the head metadata includes second resting information items, which define one or more locations on the head where the spectacle frame rests on the head.

6. The method as claimed in claim 5, wherein the first resting information items comprise a resting point associated with a nose bridge of the spectacle frame and wherein the second information items comprise a placement point associated with a nasal bridge of the 3D model of the head, wherein the approximate virtual positioning further comprises a positioning where the resting point of the frame metadata is brought into correspondence with the placement point of the head metadata.

7. The method as claimed in claim 6, further comprising: calculating an information item defining a nasal bridge of the head; and determining the placement point on the basis of the information item defining the nasal bridge, wherein the information item defining the nasal bridge is a 3D curve.

8. The method as claimed in claim 7, wherein the 3D curve has a parameterization, and wherein the parameterization is defined as a function of a coordinate of the y-direction.

9. The method as claimed in claim 1, further comprising at least one process from the group of: automatically determining at least one portion of the frame metadata and/or the head metadata; or manually determining at least one portion of the frame metadata and/or the head metadata.

10. The method as claimed in claim 9, wherein the automatic determination comprises machine learning.

11. The method as claimed in claim 1, wherein the frame metadata comprise bending information items about a flexibility of the spectacle frame, and wherein the virtual positioning comprises a bending of the spectacle frame on the basis of the information items about the flexibility.

12. The method as claimed in claim 1, further comprising: displaying the 3D model of the head with the spectacle frame positioned thereon; and changing a position of the displayed spectacle frame on the nose to an altered position, wherein the altered position of the spectacle frame on the nose is stored as a new placement point of the head metadata.

13. The method as claimed in claim 1, further comprising: recording images of the head to determine the 3D model of the head; capturing an ambient illumination when recording the images; performing an image synthesis with a virtual illumination on a basis of the captured ambient illumination; and displaying the 3D model of the head with the spectacle frame positioned thereon.

14. A computer program stored on a non-transitory computer storage medium and comprising a program code which, when executed on a processor, carries out the method as claimed in claim 1.

15. An apparatus for virtual fitting of a pair of spectacles, comprising: a processor; a display; and a computer program as claimed in claim 14 being stored on a non-transitory computer storage medium, the computer program being configured to be executed on the processor.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The disclosure will now be described with reference to the drawings wherein:

(2) FIG. 1A shows an apparatus for virtual fitting of a pair of spectacles according to one exemplary embodiment,

(3) FIG. 1B shows an example of an implementation of a camera device of FIG. 1A;

(4) FIG. 2 shows a flowchart of a method according to one exemplary embodiment;

(5) FIGS. 3A to 3C show views for elucidating 3D models of a head;

(6) FIG. 4 shows an illustration for elucidating a 3D model of a spectacle frame;

(7) FIG. 5 shows a flowchart of a method for determining frame metadata according to one exemplary embodiment;

(8) FIGS. 6A and 6B show illustrations for elucidating the manual determination of metadata;

(9) FIGS. 7A and 7B show illustrations of spectacle frames;

(10) FIG. 8 shows a flowchart of a method for determining a resting curve;

(11) FIG. 9 shows a flowchart of a method for fitting a pair of spectacles;

(12) FIGS. 10A and 10B show illustrations for elucidating steps of the method of FIG. 9: and

(13) FIG. 11 shows an illustration for elucidating a rotation of the spectacle frame.

DESCRIPTION OF EXEMPLARY EMBODIMENTS

(14) FIG. 1A shows an exemplary embodiment of an apparatus for virtual fitting of a pair of spectacles according to one exemplary embodiment. The apparatus of FIG. 1A comprises a computing device 11, a processor 12 and a memory 13. The memory 13 serves to store data and, in the exemplary embodiment of FIG. 1A, comprises a random access memory (RAM), a read-only memory (ROM) and one or more mass storage media (hard disk, solid-state disk, optical drive, etc.). A program is saved in the memory 13, said program, when executed on the processor 12, carries out virtual fitting of a pair of spectacles as described above.

(15) The apparatus of FIG. 1A further comprises a display 16 which displays a head of a person together with a spectacle frame when the computer program is executed on the processor 12. User inputs can be implemented by way of one or more input appliances 17, for example keyboard and mouse. Additionally or alternatively, the display 16 can be a touch-sensitive screen (touchscreen) in order to implement an input.

(16) The apparatus of FIG. 1A furthermore comprises an interface 14 to a network 18, by means of which data can be received. In particular, this allows 3D models of spectacle frames with texture to be received. In some exemplary embodiments, data are also transmitted to a further computing device via the interface 14 in order to carry out calculations, such as the discussed calculations of metadata, on said further computing device.

(17) In order to create a 3D model of a head of a person, to which the pair of spectacles should be fitted, the apparatus of FIG. 1A optionally comprises a camera device 15, by means of which a plurality of images of the person can be recorded from different directions; it being possible, as described, to determine the 3D model therefrom.

(18) FIG. 1B shows an exemplary embodiment for a camera device 15 of FIG. 1A. In the exemplary embodiment of FIG. 1B, a semicircular arrangement 110 of cameras is fastened to a column 19. A person can then position themselves in such a way that a head 111 of the person, as shown in FIG. 1B, is positioned in the semicircular arrangement 110 and can be recorded from different directions. From this, a 3D model of the head 111 can be created, as already explained above, wherein the texture of the model also arises from the image recordings. Moreover, the camera device of FIG. 1B comprises a panoramic camera 112 for the purposes of being able to ascertain an illumination during the image recording; this can then be used for an image synthesis, as described.

(19) FIG. 2 shows a method according to one exemplary embodiment. A 3D model of a head of a user is provided in step 20 of FIG. 2, for example on the basis of camera recordings as described above, in particular using recordings of the camera device 15 of FIG. 1A. A frame 3D model for a spectacle frame to be fitted is provided in step 21, for example by a manufacturer.

(20) Head metadata are provided on the basis of the head 3D model in step 22, as already described above, and frame metadata are provided on the basis of the frame 3D model in step 23, likewise as already described above. Steps 20 to 23 can also be carried out in a different sequence. Thus, step 22 and step 21 can be interchanged. The provision of the metadata in steps 22 and 23 can be implemented by newly calculating the metadata or by reading previously calculated metadata from a memory.

(21) Here, the metadata can be calculated and provided as discussed above.

(22) Then, the spectacle frame is positioned in virtual fashion in step 24, i.e., it is fitted to the head using the head metadata and the frame metadata. To this end, use is made of the already discussed methods. Thus, the virtual positioning comprises an approximate positioning on the basis of the head metadata and frame metadata and fine positioning on the basis of the 3D models of the head and of the spectacle frame.

(23) Then, the head with the spectacle frame is presented on a display in step 25, on the display 16 in the case of the apparatus of FIG. 1A. This then allows the user to assess the look of the spectacle frame on the head.

(24) Then, where necessary, the position of the spectacle frame can be displaced on the head in step 26 as described, for example, it can be displaced on the nose. This can be carried out by the user or else by an optician.

(25) Optionally, use can then be made of fitting data, which describe the fitting of the pair of spectacles to the head, for the purposes of adapting a real pair of spectacles. By way of example, the earpieces of the real spectacle frame can be bent like during the virtual fitting of the model.

(26) The method of FIG. 2 can be repeated for a plurality of spectacle frames in order to provide the user with an impression of the look of different spectacle frames. Then, the user can choose a spectacle frame on the basis of the presentations.

(27) Various steps of the method in FIG. 2 are now explained in more detail with reference to FIGS. 3 to 10.

(28) FIGS. 3A and 3C show an illustration for explaining a 3D model of the head. 31a denotes an example of a usable coordinate system, wherein the coordinate system—as explained above—is fixedly connected to the head, i.e., the coordinate system does not change its position and orientation relative to the head in the case of a rotation of the head. Here, FIG. 3A illustrates a 3D model 30 in the form of a triangle mesh with a plurality of vertices that are connected by edges. FIG. 3B shows a combination 31 of the triangle mesh and a texture. FIG. 3C shows a representation 32, as may be implemented on a screen on the basis of a model, in which only the texture is visible but not the individual vertices that are explicitly presented in FIGS. 3A and 3B for elucidation purposes. FIG. 4 shows a representation of the spectacle frame on the basis of the model, together with a coordinate system 41. The spectacle frame of 40 of FIG. 4 has a right spectacle earpiece 42A, a left spectacle earpiece 42B, a right hinge 43A, a left hinge 43B, a right frame rim 44A, a left frame rim 44B and a nose bridge 45

(29) FIG. 5 shows a flowchart of a method for determining frame metadata, i.e., an example for implementing step 23 in FIG. 2. Prominent points in the 3D model of the frame are determined by means of 3D feature descriptors in step 50. These prominent points are classified and filtered in step 51. As an alternative to the determination by means of 3D feature descriptors, points can also be marked manually, as elucidated in FIGS. 6A and 6B. To this end, the spectacle frame 40, which was already described with reference to FIG. 4, is presented on a display and relevant points are marked. To this end, FIGS. 6A and 6B show ear resting points 61, hinge points 62, positions of nose pads 63 and a center point 64 of the nose bridge, which may serve as a resting point, as examples. It should be noted that the resting point need not lie directly on the spectacles but may also be spaced apart from the actual frame, particularly in the case of nose pads.

(30) There is a coordinate transformation into the coordinate system of the principal axis of the prominent points classified in step 51 by a principal component analysis (PCA) in step 52 of FIG. 5 and nose bridge and hinges are localized in step 53 with the aid of the classified prominent points. Techniques to this end were likewise already described above.

(31) Then, a point for resting on the nasal bridge is determined as a resting point in step 54 for the approximate positioning, as already described above. The 3D model of the spectacle frame is segmented into components (left spectacle earpiece, right spectacle earpiece and the remaining frame) in step 55 and a resting region at the earpiece lower side in the form of a resting curve is determined in step 56, as already described above.

(32) The earpiece lower side, on which the resting curve is determined, is presented for a spectacle frame 70 in FIG. 7A and for the spectacle earpieces 71 and 72 in a corresponding magnified view in FIG. 7B as a single point representation. Calculations need not be carried out with the aid of such a great number of points by using the resting region; the resting curve can be used instead.

(33) FIG. 8 shows a flowchart of a method for determining the resting curve, i.e., a more detailed implementation of step 56 in FIG. 5. Start point and end point of the respective earpiece part (left spectacle earpiece or right spectacle earpiece) are described in step 80. Here, the method of FIG. 8 is carried out separately for the left spectacle earpiece and the right spectacle earpiece. A potential ear resting region is determined in step 81 and the resting curve is determined in step 82 by means of the “sliding window” technique described above. Here, steps 80 to 82 are carried out as already described in more detail above.

(34) FIG. 9 shows a flowchart of a method for virtual fitting and positioning of a spectacle frame on a head, wherein use is made of the models and metadata explained above. FIG. 9 represents an implementation example for steps 24 to 26 in FIG. 2.

(35) There is approximate positioning in step 90 of FIG. 9, wherein the resting point of the metadata of the spectacle frame and the placement point of the metadata of the head are brought into correspondence. Then, the frame is bent open in step 91, wherein use is made here of the cylinder intersection methods already explained above. In the case of frames in which an inclination can be modified, i.e., the inclination can be modified by rotating the earpiece about the x-axis (the coordinate system 41 of FIG. 4 is used), this is implemented in step 92. Step 92 can be skipped in the case of spectacle frames where the frame metadata indicate that such a rotation is not possible.

(36) Then, fine positioning of the frame in the xy-plane perpendicular to the z-direction is carried out in step 93, wherein the techniques already described above for the fine positioning are used here. Then, there is an image synthesis of the frame and of the head in step 94, corresponding to the positioning in steps 90 to 93, wherein, as explained above, light sources may be taken into account. Then, the user interacts with the represented model in step 95, i.e., the display of the frame and the head, using one or more input appliances (for example the input appliances 17 in FIG. 1A). As result of this, a navigation can be carried out, for example as indicated in step 96, i.e., the displayed model can be rotated, magnified or reduced in size. Then, step 94 is carried out again on the basis of these inputs; i.e., the image is redisplayed in accordance with the navigation. There may also be an input, also referred to as gesture, for rotating the frame, inter alia in order to compensate asymmetries in the face, as described above. In this case, the positioning is recalculated from step 92. Finally, the pair of spectacles can be moved up and down along the nasal bridge. This corresponds to a modification of the initial approximate positioning and hence the method is carried out again from step 90 in this case, wherein the newly set position on the nasal bridge is used as approximate positioning in step 90.

(37) FIGS. 10A and 10B show representations of a head 100 together with a spectacle frame 102 for elucidating method steps of FIG. 9. FIG. 10A elucidates how the spectacle frame 102 can be displaced upward and downward along the nasal bridge 101 of the head 100 in accordance with arrows 103. This is an example of the movement on the nasal bridge according to step 96. FIG. 10B elucidates a bending open of the earpieces of the spectacle frame 102 in accordance with arrows 104, as carried out in step 91 of FIG. 9.

(38) FIG. 11 shows an illustration for explaining the rotation of the frame about the x-axis by means of a cylinder intersection, as already explained above. FIG. 11 shows a 3D model of a head 110 and a 3D model of a spectacle frame 111, in each case as a 3D mesh. During the approximate positioning, the frame was initially positioned on the basis of a resting point 116. An axis 115 extends through the resting point 116 in the x-direction. A cylinder, indicated by circles 112, 113, has a radius to an ear resting point of the 3D model of the head 110. The point of intersection of the cylinder with a resting region of the spectacle frame, which is represented by the 3D model of the spectacle frame 111, with the cylinder yields a direction for the spectacle earpieces of the spectacle frame and consequently yields an angle through which the spectacle frame should be rotated about the axis 115. In the present case, a rotation is carried out from a position where the spectacle earpieces extend in the direction of lines 119A, 119B such that the spectacle earpieces now extend in the direction of the lines 1110A, 1110B.

(39) Using the described methods and apparatuses, it is consequently possible to undertake virtual accurate fitting of a spectacle frame to a head, which then optionally, according to step 27 in FIG. 2, can also be used for fitting a real spectacle frame.

(40) At least some possible exemplary embodiments are specified in the clauses below:

(41) Clause 1. A computer-implemented method for virtual fitting of a pair of spectacles, comprising: i) fine virtual positioning of the spectacle frame on the basis of a 3D model (30, 31) of a head and a 3D model (40) of a spectacle frame, ii) characterized by iii) approximate virtual positioning of the spectacle frame on the head on the basis of head metadata for the 3D model (30, 31) of the head and frame metadata for the 3D model of the spectacle frame before the fine virtual positioning.

(42) Clause 2. The method according to clause 1, characterized in that the frame metadata comprise first resting information items, which define one or more locations on the spectacle frame where the spectacle frame rests on the head, and/or the head metadata comprise second resting information items, which define one or more locations on the head where the spectacle frame rests on the head.

(43) Clause 3. The method according to clause 2, characterized in that the first resting information items comprise a resting point associated with a nose bridge of the spectacle frame and wherein the second information items comprise a placement point associated with a nasal bridge of the 3D model of the head, and in that the approximate virtual positioning comprises a positioning where the resting point of the frame metadata is brought into correspondence with the placement point of the head metadata.

(44) Clause 4. The method according to clause 3, characterized by calculating information items characterizing a nasal bridge of the head and determining the placement point on the basis of the information items characterizing the nasal bridge.

(45) Clause 5. The method according to any one of clauses 1-4, characterized in that if a horizontal direction is defined corresponding to a connecting line of the eyes of the head when the head is held erect, if a vertical direction is defined as y-direction and if a direction perpendicular to the x- and y-direction is defined as z-direction, then a displacement of the spectacle frame in the z-direction is no more than 10% of a displacement of the spectacle frame in the x-y-direction during fine virtual positioning.

(46) Clause 6. The method according to any one of clauses 1-5, characterized in that the method comprises at least one process from the group automatically determining at least one portion of the frame metadata and/or the head metadata manually determining at least one portion of the frame metadata and/or the head metadata.

(47) Clause 7. The method according to clause 6, characterized in that the automatic determination comprises machine learning.

(48) Clause 8. The method according to any one of clauses 1-7, characterized in that the method for providing the frame metadata comprises identifying prominent points in the 3D model of the spectacle frame and/or classifying prominent points in the 3D model of the spectacle frame.

(49) Clause 9. The method according to clause 8, characterized in that the method further comprises an application of a coordinate transformation on the basis of the identified prominent points to at least one of the group of: the 3D model of the spectacle frame the identified prominent points the frame metadata.

(50) Clause 10. The method according to any one of clauses 1-9, characterized in that the provision of the frame metadata comprises a segmentation of the 3D model of the spectacle frame into components of the spectacle frame, wherein the components preferably comprise at least one component of the group earpiece parts a remaining part of the spectacle frame apart from earpiece parts and/or wherein the head metadata comprise resting regions for the ears of the head, which regions are calculated on the basis of the segmented earpiece parts.

(51) Clause 11. The method according to any one of clauses 1-10, characterized in that the frame metadata comprise bending information items about a flexibility of the spectacle frame and in that the virtual positioning comprises a bending of the spectacle frame on the basis of the information items about the flexibility.

(52) Clause 12. The method according to any one of clauses 1-11, characterized in that the method comprises changing the position of the spectacle frame on the nose following the display, wherein an altered position of the spectacle frame on the nose is stored as a new placement point of the head metadata.

(53) Clause 13. The method according to any one of clauses 1-12, characterized in that the method further comprises determining the 3D model of a head on the basis of image recordings and determining an ambient illumination when the images were recorded, and in that the display of the head with the frame positioned thereon comprises an image synthesis with a virtual illumination on the basis of the captured ambient illumination.

(54) Clause 14. A computer program comprising a program code which, when executed on a processor, carries out the method according to any one of clauses 1-13.

(55) Clause 15. An apparatus for virtual fitting of a pair of spectacles, comprising: a processor and a display, characterized by a computer program according to clause 14 to be executed on the processor.

(56) The foregoing description of the exemplary embodiments of the disclosure illustrates and describes the present invention. Additionally, the disclosure shows and describes only the exemplary embodiments but, as mentioned above, it is to be understood that the disclosure is capable of use in various other combinations, modifications, and environments and is capable of changes or modifications within the scope of the concept as expressed herein, commensurate with the above teachings and/or the skill or knowledge of the relevant art.

(57) The term “comprising” (and its grammatical variations) as used herein is used in the inclusive sense of “having” or “including” and not in the exclusive sense of “consisting only of.” The terms “a” and “the” as used herein are understood to encompass the plural as well as the singular.

(58) All publications, patents, and patent applications cited in this specification are herein incorporated by reference, and for any and all purposes, as if each individual publication, patent or patent application were specifically and individually indicated to be incorporated by reference. In the case of inconsistencies, the present disclosure will prevail.