Methods and systems for detecting cracks in illuminated electronic device screens

1. A method performed by one or more computing devices for detecting cracks in a screen of an electronic device, the method comprising:

overlaying at least one kernel on an image of the screen of the electronic device at a first selected location, the kernel comprising a matrix having a line region and a non-line region;

determining a first average brightness value for the line region and a second average brightness value for the non-line region at the first selected location;

automatically identifying a crack at the first selected location if the first average brightness value is greater than the second average brightness value by at least a predetermined threshold amount;

overlaying the at least one kernel on the image of the screen of the electronic device at a second selected location;

determining a third average brightness value for the line region and a fourth average brightness value for the non-line region at the second selected location; and

automatically identifying a crack at the second selected location if the third average brightness value is greater than the fourth average brightness value by at least the predetermined threshold amount.

2. The method of claim 1 , further comprising:

at the first location, generating an image pyramid with multiple layers of the image, each of the layers being down sampled to a different degree; and

for each of the layers, overlaying the kernel on the layer to obtain, at various locations of the layer, average brightness values for each of the line region and the non-line region.

3. The method of claim 1 wherein the at least one kernel includes at least a first kernel and a second kernel, wherein the line region in the first kernel has a first orientation, and wherein the line region in the second kernel has a second orientation different from the first orientation.

4. The method of claim 1 wherein the line region comprises a first line region portion and and a second line region portion, and wherein the method further comprises:

at the first location, determining brightness values for each of the first line region portion and the second line region portion, wherein the first average brightness value for the line region is an average of the brightness value for the first line region portion and the brightness value for second line region portion.

5. The method of claim 1 wherein the the non-line region comprises a first non-line region portion and a second non-line region portion, and wherein the method further comprises:

at the first location, determining brightness values for each of the first non-line region portion and the second non-line region portion, wherein the second average brightness value for the non-line region is an average of the brightness value for the first non-line region portion and the brightness value for the second non-line region portion.

6. The method of claim 1 wherein the the non-line region comprises a first non-line region portion positioned on a first side of the line region and a second non-line region portion positioned on a second side of the line region opposite the first side.

7. The method of claim 1 wherein the image of the screen is a processed image, and wherein the method further comprises:

forming the processed image, wherein forming the processed image includes:

obtaining a lighted image of the screen;

obtaining a non-lighted image of the screen; and

subtracting the lighted image from the non-lighted image.

8. A method performed by one or more computing devices for detecting cracks in a screen of an electronic device, the method comprising:

overlaying at least one kernel on an image of the screen of the electronic device at a selected location, the kernel comprising a matrix having a line region and a non-line region, wherein the line region comprises a first line region portion and a second line region portion;

determining brightness values for each of the first line region portion and the second line region portion;

determining a first average brightness value for the line region and a second average brightness value for the non-line region, wherein the first average brightness value for the line region is an average of the brightness value for the first line region portion and the brightness value for second line region portion;

automatically identifying a crack at the selected location if the first average brightness value is greater than the second average brightness value by at least a predetermined threshold amount; and

performing a secondary check on the identified crack to detect a false-positive crack identification.

9. The method of claim 8 , wherein the secondary check includes determining whether at least one of the following conditions is satisfied:

(a) a brightness value of a first non-line region portion and a brightness value of a second non-line region portion are within a predetermined range of one another;

(b) the brightness value of the first non-line region portion or the brightness value of the second non-line region portion is below a predetermined threshold level;

(c) the first average brightness value of the line region is greater than the second average brightness value of the non-line region by a predetermined threshold amount;

(d) any brightness value in the first line region portion, the second line region portion, the first non-line region portion, or the second non-line region portion is greater than a predetermined threshold value;

(e) any brightness value in the first line region portion, the second line region portion, the first non-line region portion, or the second non-line region portion is lower than a predetermined threshold value; and

(f) any three consecutive brightness values in the line region do not increase or decrease monotonically,

wherein, if any of conditions (a)-(f) are satisfied, then the identified crack is discarded.

10. The method of claim 1 wherein the at least one kernel is a plurality of kernels, and wherein if a crack is identified at the first selected location for any one of the plurality of kernels, any remaining kernels in the plurality of kernels are not overlayed on the image of the screen of the electronic device at the first selected location.

11. A computing system for detecting a crack in an illuminated screen of a mobile phone, the computing system comprising:

one or more processors for executing computer-executable instructions; and

a non-transitory computer-readable storage medium storing computer-executable instructions for execution by the one or more processors, wherein execution of the computer-executable instructions cause the one or more processors to:

obtain an image of the illuminated screen of the mobile phone;

overlay at least one kernel on a screen region of the image at a selected location of the image, the kernel comprising a matrix having a first region and a second region;

determine a first average brightness value for the first region and a second average brightness value for the second region; and

automatically identify a crack at the selected location if the first average brightness value is greater than the second average brightness value by at least a predetermined threshold amount.

12. The computing system of claim 11 wherein execution of the computer-executable instructions further cause the processor to:

generating multiple copies of the image, each of the copies being down sampled to a different degree; and

for each of the copies, overlay the at least one kernel on the copy to obtain the first average brightness value and the second average brightness value for the first and second regions.

13. The computing system of claim 11 wherein execution of the computer-executable instructions further cause the processor to:

before overlaying the at least one kernel on the screen region, identify cracks in the mobile phone outside of the screen region of the image, wherein, if one or more cracks outside of the screen region are identified, the mobile phone is identified as damaged.

14. The computing system of claim 13 wherein checking for cracks on the mobile phone outside of the screen region of the image includes using at least one of a Canny edge detector and a Hough transform.

15. The computing system of claim 11 wherein execution of the computer-executable instructions further cause the processor to perform a secondary inspection of the identified crack to detect a false-positive crack identification.

16. The computing system of claim 11 wherein the at least one kernel is a plurality of kernels, and wherein if a crack is identified at the selected location for any one of the plurality of kernels the computer-executable instructions further control the one or more computing systems to not overlay any additional kernels on the screen region of the image at the selected location.

17. The computing system of claim 11 wherein the image is a processed image, and wherein execution of the computer-executable instructions further cause the processor to:

generate the processed image, wherein generating the processed image includes:

obtaining a lighted image of the screen;

obtaining a non-lighted image of the screen; and

subtracting the lighted image from the non-lighted image.

18. A computing system for detecting a crack in a screen of an electronic device, the computing system comprising:

one or more processors for executing computer-executable instructions; and

a non-transitory computer-readable storage medium storing computer-executable instructions for execution by the one or more processors, wherein execution of the computer-executable instructions cause the one or more processors to:

obtain an image of the electronic device;

overlay at least one kernel on a screen region of the image at a selected location of the image, the kernel comprising a matrix having a first region and a second region;

determine a first average brightness value for the first region and a second average brightness value for the second region, wherein the second region comprises a first portion positioned on a first side of the first region and a second portion positioned on a second side of the first region opposite the first side; and

automatically identify a crack at the selected location if the first average brightness value is greater than the second average brightness value by at least a predetermined threshold amount.

19. The computing system of claim 18 wherein execution of the computer-executable instructions further cause the one or more processors to perform a secondary inspection of the identified crack to detect a false-positive crack identification.