Electronic devices for defect detection

1. An electronic device configured for defect detection, comprising:

a processor;

memory in electronic communication with the processor;

instructions stored in the memory, the instructions being executable to:

obtain image data;

perform background suppression on the image data based on a transform of the image data to obtain a score map;

apply thresholding to the score map to generate a detection mask, wherein the thresholding comprises bi-thresholding;

detect any defects based on the detection mask; and

indicate any defects.

2. The electronic device of claim 1 , wherein applying bi-thresholding comprises:

selecting any pixel with a score that exceeds a first threshold, wherein any pixel with a score that exceeds the first threshold is a seed;

selecting any pixel with a score that exceeds a second threshold and that is located within a predetermined proximity to the seed, wherein the selection is based on a forward scan; and

selecting any pixel with a score that exceeds the second threshold and that is located within the predetermined proximity to the seed, wherein the selection is based on a backward scan.

3. The electronic device of claim 1 , wherein applying bi-thresholding comprises locally adaptive thresholding.

4. The electronic device of claim 1 , wherein the transform comprises a two-dimensional discrete transform.

5. The electronic device of claim 4 , wherein the transform is one of the group consisting of a Discrete Fourier Transform, Walsh-Hadamard Transform, a transform based on singular value decomposition and a transform based on discrete polynomials.

6. The electronic device of claim 1 , wherein the transform comprises a one-dimensional discrete transform.

7. The electronic device of claim 6 , wherein the transform is one of the group consisting of a Discrete Fourier Transform and a Walsh-Hadamard Transform.

8. The electronic device of claim 6 , wherein the one-dimensional discrete transform is performed on each row of the image data.

9. The electronic device of claim 6 , wherein the one-dimensional discrete transform is performed on each column of the image data.

10. The electronic device of claim 1 , wherein the transform comprises at least one one-dimensional transform in combination with a two-dimensional transform.

11. The electronic device of claim 1 , wherein detecting any defects comprises determining if the detection mask includes a line.

12. The electronic device of claim 11 , wherein determining if the detection mask includes a line comprises:

grouping any adjacent selected pixels from the detection mask to obtain one or more pixel blobs;

sorting the one or more pixel blobs in descending order of area;

determining a major axis and a minor axis of each pixel blob;

selecting a pixel blob with a largest area and with a line ratio that exceeds a line threshold to obtain a selected pixel blob;

selecting any additional pixel blob within a region;

combining the selected pixel blob and any additional pixel blob to obtain a combined pixel blob area;

determining a ratio of the combined pixel blob area over an area of unselected pixels in the detection mask; and

detecting the combined pixel blob area as a line if the ratio is above a threshold.

13. The electronic device of claim 1 , wherein performing background suppression on the image data comprises:

obtaining reference image data;

performing a forward transform on the image data to obtain an image data coefficient;

performing the forward transform on the reference image data to obtain a reference image data coefficient;

extracting information from the image data coefficient based on the reference image data coefficient to obtain extracted information;

performing an inverse transform on the extracted information to obtain a feature map; and

generating the score map based on the feature map.

14. The electronic device of claim 13 , wherein the extracted information comprises one of sign information and phase information.

15. The electronic device of claim 13 , wherein the forward transform is the transform based on singular value decomposition.

16. The electronic device of claim 13 , wherein the forward transform is a Walsh-Hadamard Transform.

17. A method for defect detection on an electronic device, comprising:

obtaining image data;

performing, by the electronic device, background suppression on the image data based on a transform of the image data to obtain a score map;

applying, by the electronic device, thresholding to the score map to generate a detection mask, wherein the thresholding comprises bi-thresholding;

detecting any defects based on the detection mask; and

indicating any defects.

18. The method of claim 17 , wherein the bi-thresholding comprises:

selecting any pixel with a score that exceeds a first threshold, wherein any pixel with a score that exceeds the first threshold is a seed;

selecting any pixel with a score that exceeds a second threshold and that is located within a predetermined proximity to the seed, wherein the selection is based on a forward scan; and

selecting any pixel with a score that exceeds the second threshold and that is located within the predetermined proximity to the seed, wherein the selection is based on a backward scan.

19. The method of claim 17 , wherein the bi-thresholding comprises locally adaptive thresholding.

20. The method of claim 17 , wherein the transform comprises a two-dimensional discrete transform.

21. The method of claim 20 , wherein the transform is one of the group consisting of a Discrete Fourier Transform, Walsh-Hadamard Transform, a transform based on singular value decomposition and a transform based on discrete polynomials.

22. The method of claim 17 , wherein the transform comprises a one-dimensional discrete transform.

23. The method of claim 22 , wherein the transform is one of the group consisting of a Discrete Fourier Transform and Walsh-Hadamard Transform.

24. The method of claim 22 , wherein the one-dimensional discrete transform is performed on each row of the image data.

25. The method of claim 22 , wherein the one-dimensional discrete transform is performed on each column of the image data.

26. The method of claim 17 , wherein the transform comprises at least one one-dimensional transform in combination with a two-dimensional transform.

27. The method of claim 17 , wherein detecting any defects comprises determining if the detection mask includes a line.

28. The method of claim 27 , wherein determining if the detection mask includes a line comprises:

grouping any adjacent selected pixels from the detection mask to obtain one or more pixel blobs;

sorting the one or more pixel blobs in descending order of area;

determining a major axis and a minor axis of each pixel blob;

selecting a pixel blob with a largest area and with a line ratio that exceeds a line threshold to obtain a selected pixel blob;

selecting any additional pixel blob within a region;

combining the selected pixel blob and any additional pixel blob to obtain a combined pixel blob area;

determining a ratio of the combined pixel blob area over an area of unselected pixels in the detection mask; and

detecting the combined pixel blob area as a line if the ratio is above a threshold.

29. The method of claim 17 , wherein performing background suppression on the image data comprises:

obtaining reference image data;

performing a forward transform on the image data to obtain an image data coefficient;

performing the forward transform on the reference image data to obtain a reference image data coefficient;

extracting information from the image data coefficient based on the reference image data coefficient to obtain extracted information;

performing an inverse transform on the extracted information to obtain a feature map; and

generating the score map based on the feature map.

30. The method of claim 29 , wherein the extracted information comprises one of sign information and phase information.

31. The method of claim 29 , wherein the forward transform is the transform based on singular value decomposition.

32. The method of claim 29 , wherein the forward transform is a Walsh-Hadamard Transform.

33. An electronic device configured for defect detection, comprising:

a processor;

memory in electronic communication with the processor;

instructions stored in the memory, the instructions being executable to:

obtain image data;

perform background suppression on the image data based on a Walsh-Hadamard Transform of the image data to obtain a score map;

apply thresholding to the score map to generate a detection mask;

detect any defects based on the detection mask; and

indicate any defects.

34. A method for defect detection on an electronic device, comprising:

obtaining image data;

performing, by the electronic device, background suppression on the image data based on a Walsh-Hadamard Transform of the image data to obtain a score map;

applying, by the electronic device, thresholding to the score map to generate a detection mask;

detecting any defects based on the detection mask; and

indicating any defects.