# US 6,508,553 B2 · Interactive eyewear selection system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 6,508,553 B2 - Document ID: US6508553B2 - Status: Granted - Kind: B2 - Application number: 09/854,861 - Filed: 2001-05-14 - Published: 2001-10-04 - Granted: 2003-01-21 - Earliest claimed: 1998-09-22 - Assignees: Virtual Visual Devices, LLC - Inventors: Feng Gao; Wei Li - Web page: https://patentgrep.com/patent/US6508553B2 ## CPC - G02C 13/005: {Measuring geometric parameters required to locate ophtalmic lenses in spectacles frames (apparatus for testing or instruments for examining the eyes per se A61B3/00; machines or devices and accessories for grinding the edges of lenses using spectacles as a template B24B9/144)} - G02C 13/003: {Measuring during assembly or fitting of spectacles} - G02C: SPECTACLES; SUNGLASSES OR GOGGLES INSOFAR AS THEY HAVE THE SAME FEATURES AS SPECTACLES; CONTACT LENSES (trial frames for testing the eyes A61B3/04; goggles or eyeshields not having the same features as spectacles A61F9/00) ## Abstract A method of determining eyeglass frame parameters from a digital image of the person's face and a digital image of an eyeglass frame. The method comprises the steps of: (a) providing a digital image of the person's face; (b) determining a physical scale of the face using a scaling factor; (c) locating center points P1, P2 of the left and right irises in the face image, and determining a pupil midpoint Pc; (d) providing a digital image of an eyeglass frame; (e) adjusting, if necessary, the size of the frame to the physical scale of the face; (f) combining the frame and face images such that the frame is superimposed on the face; (g) aligning the frame and face in accordance with the frame center point fc and the pupil midpoint Pc; and (h) determining the frame parameters. ## Claims ### Claim 1 A method of automatically determining eyeglass frame parameters from digital image of a person's face and a digital image of an eyeglass frame, said method comprising the steps of: (a) providing a digital image of the person's face, the face depicted in the face image containing left and right eyes with left and right irises, respectively; (b) determining a physical scale of the face in the face image; (c) locating a center point P1 of the left iris of the face image, locating a center point P2 of the right iris of the face image, and determining a pupil midpoint Pc between the center points P1 and P2; (d) providing a digital image of an eyeglass frame, the frame depicted in the digital image having left and right frame rims; (e) adjusting, if necessary, the size of the frame of the frame image to the physical scale of the face of the face image; (f) combining the frame image with the face image to create a combined image, the combined image depicting the frame of the frame image superimposed on the face of the face image; (g) aligning the frame relative to the face in the combined image; and (h) determining the left and the right monocular PDs from the center points P1 and P2 and from the pupil midpoint Pc. ### Claim 2 (depends on 1) The method of claim 1, wherein the frame of the frame image imcludes a frame reference point located between the frame rims, and wherein said step (g) includes aligning the frame with the face in accordance with the frame reference point. ### Claim 3 (depends on 2) The method of claim 2, wherein the frame reference point is the midpoint fc between the frame rims, and wherein said step (g) includes aligning the frame with the face in accordance with the midpoint fc and the pupil midpoint Pc. ### Claim 4 (depends on 1) The method of claim 1, wherein the left and the right frame rims of the frame image contain first and second points, respectively, and wherein a first frame parameter d1 is defined as the distance from the first point to the bottom of the left frame rim and wherein a second frame parameter d2 is defined as the distance from the second point to the bottom of the right frame rim, said method further comprising the steps of: (i) determining the left and the right ocular centers from the first and the second frame parameters, d1 and d2, respectively. ### Claim 5 (depends on 4) The method of claim 4, wherein the first and the second points are the centroids of the left and the right rims, respectively. ### Claim 6 (depends on 4) The method of claim 4, further comprising the steps of: (j) determining the radii of the left and the right irises of the face image; and (k) determining the left and the right seg-heights from the radii of the left and the right irises, respectively, and from the first and the second frame parameters, d1 and d2, respectively. ### Claim 7 (depends on 6) The method of claim 6, further comprising the steps of: (l) adjusting the position of the frame relative to the position of the face in the combined image by an adjustment factor, the monocular PDs, ocular centers, and seg-heights being adjusted in accordance with the adjustment factor. ### Claim 8 (depends on 7) The method of claim 7, wherein said step (1) includes automatically adjusting the position of the frame relative to the position of the face. ### Claim 9 (depends on 7) The method of claim 7, wherein said step (1) includes manually adjusting the position of the frame relative to the position of the face. ### Claim 10 A method of processing a digital eyeglass frame image which depicts an eyeglass frame having a nosepad, said method comprising the steps of: (a) providing a first digital color image depicting the eyeglass frame in front of a first background; (b) providing a second digital color image, the second image depicting a second background without the eyeglass frame; (c) generating first and second gray level images from the first and the second digital color images, respectively, the first and the second gray level images each containing gray level pixel data, the pixel data of the first gray level image including eyeglass frame data and background data, the eyeglass frame data including nosepad data; (d) determining the locations of the eyeglass frame data in the first gray level image by determining the difference between the pixel data of the first gray level image and the pixel data of the second gray level image and comparing the difference with a first threshold value; (e) determining the locations of the nosepad data in the first gray level image by comparing the eyeglass frame data of the first gray level image to a second threshold value; and (f) modifying the first digital color image by assigning a pre-determined color N to locations of the first color image that correspond to the locations of the nosepad data determined in said step. ### Claim 11 (depends on 10) The method of claim 10, wherein the first and the second backgrounds are the same pre-set background. ### Claim 12 (depends on 10) The method of claim 10, wherein the first gray level image is a sub-image of a gray level image of the first digital color image. ### Claim 13 (depends on 10) The method of claim 10, wherein the locations of the eyeglass frame data are those locations wherein the difference between the pixel data of the first and the second gray level images is greater than the first threshold value. ### Claim 14 (depends on 13) The method of claim 13, wherein the first threshold value is 20. ### Claim 15 (depends on 10) The method of claim 10, wherein the locations of the nosepad data are those locations where the eyeglass frame data is greater than the second thershold value. ### Claim 16 (depends on 15) The method of claim 15, wherein the second threshold value is 190. ### Claim 17 (depends on 10) The method of claim 10, further comprising the steps of: (g) providing a third digital color image, the third image depicting a person's face; (h) combining the first digital color image, as modified in said step (f), with the third image to create a combined color image, the combined image depicting the eyeglass frame superimposed naturally on the person's face; and (i) mixing the pre-determined color N with the color of the person's face in the combined image where the nosepad of the eyeglass frame is depicted, such that the nosepad in the combined image appears at least partially transparent. ### Claim 18 (depends on 17) The method of claim 17, wherein the pre-determined color N is mixed with color of the face in accordance with the equation: nosepad=0.8·face color +0.2·N. ### Claim 19 A method of producing a blending coefficient for representing the degree of transparency in a transparent or semi-transparent eyeglass frame depicted in a digital image, said method comprising the steps of: (a) providing a first digital color image, the first image depicting the eyeglass frame in front of a first background; (b) providing a second digital color image, the second image depicting a second background without the eyeglass frame; (c) generating first and second gray level images from the first and the second digital color images, respectively, the first and the second gray level images each containing gray level pixel data, the pixel data of the first gray level image including eyeglass frame data and background data; (d) determining the difference between the pixel data of the first gray level image and the pixel data of the second gray level image; and (e) determining the blending coefficient from the difference determined in said step (d). ### Claim 20 (depends on 19) The method of claim 19, wherein said step (d) includes determining the standard deviation between the pixel data of the first gray level image and the pixel data of the second gray level image; and wherein said step (e) includes determining the blending coefficient from said standard deviation. ### Claim 21 (depends on 19) The method of claim 19, further comprising the step of: (f) smoothing the second digital color image before generating the second gray level image therefrom. ### Claim 22 (depends on 19) The method of claim 19, further comprising the step of: (f) normalizing the blending coefficient. ### Claim 23 (depends on 22) The method of claim 22, further comprising the step of: (g) storing the normalized blending coefficient in association with the first digital color image. ### Claim 24 A method of producing a composite digital color image of a transparent or semi-transparent eyeglass frame superimposed on a person's face, using a blending coefficient, said method comprising the steps of: (a) providing a first digital color image, the first image depicting a transparent or semi-transparent eyeglass frame; (b) providing a blending coefficient associated with the first image; (c) providing a second digital color image, the second image depicting a person's face; (d) combining the first image with the second image to create a composite digital color image, the composite image depicting the eyeglass frame superimposed naturally on the person's face; and (e) in the composite image, blending the person's face with the eyeglass frame in accordance with the blending coefficient. ### Claim 25 (depends on 24) The method of claim 24, wherein the blending coefficient is normalized. ### Claim 26 (depends on 25) The method of claim 25, wherein said steps (d) and (e) are performed in accordance with the following equation: C(x,y)=[{overscore (c(x,y))}/255]·face (x,y) +]1−{overscore (c(x,y))}/255]·frame (x,y), where C(x,y) represents the composite image, {overscore (c(x,y))} represents the normalized blending coefficient, face (x,y) represents the second image, and frame (x,y) represents the first image. ### Claim 27 A method of re-sizing an eyeglass frame depicted in a digital eyeglass frame image, said method comprising the steps of: (a) providing a digital image of an eyeglass frame, the frame having a pair of rims and a bridge, the frame having a size Q2-Q1 and the bridge having a size B2-B1; (b) employing a rim scaling factor which is a function of (i) the frame size Q2-Q1; (ii) a selected rim size EYE, and (iii) a selected bridge size DBL; (c) employing a bridge scaling factor which is a function of the bridge size B2-B1 and the selected bridge size DBL; and (d) changing the size of the frame of the digital image by scaling the pair of rims and the bridge of the frame separately, if necessary, using the rim and the bridge scaling factors. ### Claim 28 (depends on 27) The method of claim 27, wherein the rim scaling factor is determined by the following equation: (Q2-Q1)/(2·EYE+DBL), and the bridge scaling factor is determined by the following equation: (B2-B1)/DBL. ### Claim 29 A method of stimulating color contact lenses on the eyes of a person depicted in a digital color image, said method comprising the steps of: (a) providing a digital color image of a person's face, the face having a pair of eyes each containing an iris and a pupil; (b) defining an outside area on each of the pair of eyes, the outside area surrounding the iris and having a boundary; (c) defining an inside area on each of the pair of eyes, the inside area being located inside the outside area and having a boundary which substantially bounds the pupil; (d) selecting a contact lens color; and (e) creating a composite digital color image of the person's face and the contact lens color, by mixing the contact lens color with the person's face within the boundary of the outside area, the proportion of the contact lens color to the person's face varying from a higher number at the boundary of the inside area to a lower number at the boundary of the outside area, whereby the contact lens color in the composite image simulates color contact lenses being worn naturally on the eyes of the person. ### Claim 30 (depends on 29) The method of claim 29, wherein the proportion of the contact lens color to the person's face is substantially constant within the boundary of the inside area. ### Claim 31 (depends on 29) The method of claim 29, wherein the composite image is created by the following equation: Composite Image=Coef·face+(1-Coef)·lens color, where Coef is a mixing coefficient. ### Claim 32 (depends on 31) The method of claim 31, wherein the outside area has a radius r2 and the inside area has a radius r1, and wherein the mixing coefficient Coef is defined by the following equation: Coef=0.9; r