Method and system of processing an image containing undesirable pixels

• US 8,023,766 B1
• Filed: 04/30/2007
• Issued: 09/20/2011
• Est. Priority Date: 04/30/2007
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Bergman, Ruth, Ruckenstein, Gitit, Nachlieli, Hila
• Primary Examiner(s): Cunningham; Gregory F
• Application Number:
  US11/799,070
• Time in Patent Office: 1,604 Days
• Field of Search: 348/666, 358/3.26, 382/275
• US Class Current: 382/275
• CPC Class Codes: G06T 5/77 Retouching; Inpainting; Scratch removal
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06V 10/26 Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion

Abstract

In a method of processing an image containing undesirable pixels, a coarse identification of a location of the undesirable pixels is received. The coarse identification includes identification of at least one undesirable pixel and at least one desirable pixel in the image. An area in the image to be analyzed for undesirable pixel values is automatically determined according to the coarse identification received. In addition, in the area determined to be analyzed, the pixels are automatically classified as one of undesirable and desirable and the classifications of the pixels are stored.

Claims

1. A method of processing an image containing undesirable pixels, said method comprising:

receiving a coarse identification of a location of the undesirable pixels, wherein the coarse identification includes identification of at least one undesirable pixel and at least one desirable pixel in the image;

automatically determining, by a computer, an area in the image to be analyzed for undesirable pixel values according to the coarse identification received based upon a shape of the undesirable pixels being identified;

in the area determined to be analyzed, automatically classifying the pixels as one of undesirable and desirable; and

storing the classifications of the pixels.

2. The method according to claim 1, wherein automatically determining an area in the image to be analyzed further comprises determining a predetermined area.

3. The method according to claim 1, further comprising:

automatically repairing the pixels determined to be undesirable by replacing values of the undesirable pixels with replacement pixel values derived from a pixel value of at least one of the pixels determined to be desirable.

4. The method according to claim 1, further comprising:

identifying a local neighborhood of pixels around an undesirable pixel;

identifying secondary neighborhoods of pixels;

calculating feature values of the pixels contained in the local neighborhood and each of the secondary neighborhoods;

classifying the secondary neighborhoods having feature values within a predetermined range of the local neighborhood feature value as candidates for pixel value replacement;

determining a replacement pixel value from at least one of the secondary neighborhoods classified as a candidate for pixel value replacement; and

replacing the undesirable pixel value with the replacement pixel value.

5. The method according to claim 4, wherein calculating a local neighborhood feature value further comprises calculating textural measures of the local neighborhood and wherein calculating feature values of the secondary neighborhoods further comprises calculating textural measures of the secondary neighborhoods, the method further comprising:

preprocessing the secondary neighborhoods according to their respective textural measures prior to classifying the secondary neighborhoods, wherein preprocessing the secondary neighborhoods further comprises,

arranging the secondary neighborhoods according to a first textural measure; and

organizing the arranged secondary neighborhoods according to a second textural measure.

6. The method according to claim 1, wherein the undesirable pixels comprise a digital reproduction of a tear in a source image, said digital reproduction of the tear having light defective pixels and dark defective pixels, said method further comprising:

automatically determining which of the defective pixels are light defective pixels and which of the defective pixels are dark defective pixels.

7. The method according to claim 6, further comprising:

automatically repairing the light defective pixels prior to repairing the dark defective pixels.

8. The method according to claim 1, wherein receiving the coarse identification further comprises receiving a coarse identification that substantially encompasses the relatively large group of undesirable pixels.

9. The method according to claim 1, wherein receiving the coarse identification further comprises receiving a coarse identification that is substantially smaller in size than the relatively large group of undesirable pixels.

10. The method according to claim 1, wherein the undesirable pixels comprise a digital reproduction of a mole in a source image, and wherein automatically classifying the pixels as one of undesirable and desirable further comprises:

automatically determining which of the undesirable pixels form the digital reproduction of the mole by searching the analyzed area for pixel values that are relatively darker than surrounding pixel values and have at least one of a substantially round and oval shape.

11. The method according to claim 1, wherein the undesirable pixels comprise at least one of a time stamp and a date stamp on the image, and wherein automatically classifying the pixels as one of undesirable and desirable further comprises:

automatically determining which of the undesirable pixels form the at least one of the time stamp and the date stamp by searching the analyzed area for pixel values that together form distinguishable text and have color values that substantially differ from surrounding pixel color values.

12. The method according to claim 1, wherein automatically classifying the pixels further comprises automatically classifying at least one pixel as neither undesirable or desirable, said method further comprising:

determining whether the at least one pixel is most similar to pixels classified as undesirable or to pixels classified as desirable; and

classifying the at least one pixel as undesirable if the at least one pixel is most similar to the undesirable pixels and as desirable if the at least one pixel is most similar to the desirable pixels.

13. The method according to claim 1, wherein automatically classifying the pixels further comprises:

providing plural representations of the image, said plural representations comprising a gray representation and a detail-less representation; and

comparing the gray representation and the detail-less representation to determine locations of the undesirable pixels.

14. An imaging system comprising:

a detection unit;

a data storage; and

a computerized controller configured to receive a coarse identification of a location of an undesirable object in an image, said coarse identification including identification of at least one undesirable pixel and at least one desirable pixel in the image, wherein the controller is further configured to at least one of invoke and implement the detection unit to determine an area in the image to be analyzed for undesirable pixels according to the received coarse identification and an identification of a shape of the undesirable object to be located, to classify the pixels contained in the area to be analyzed as one of undesirable and desirable, and to store the pixel classifications in the data storage.

15. The imaging system according to claim 14, further comprising:

an image editing unit, wherein the controller is configured to at least one of invoke and implement the image editing unit to repair the pixels determined to be undesirable by replacing values of the undesirable pixels with replacement pixel values derived from a pixel value of at least one of the pixels determined to be desirable.

16. The imaging system according to claim 15, wherein the undesirable object comprises a digital reproduction of a tear in a source image, said digital reproduction of the tear having light defective pixels and dark defective pixels, wherein the detection unit is further configured to determine which of the undesirable pixels are light defective pixels and which of the defective pixels are dark defective pixels, and wherein the image editing unit is further configured to repair the light defective pixels prior to repairing the dark defective pixels.

17. The imaging system according to claim 15, wherein the undesirable object comprises a digital image of a mole, wherein the detection unit is further configured to determine which of the undesirable pixels form the digital reproduction of the mole by searching the analyzed area for pixel values that are relatively darker than surrounding pixel values and have at least one of a substantially round and oval shape, and wherein the image editing unit is further configured to replace values of the undesirable pixels to remove the digital reproduction of the mole.

18. The imaging system according to claim 15, wherein the undesirable object comprises at least one of a time stamp and a date stamp on the image, wherein the defect detection unit is further configured to determine which of the undesirable pixels form the at least one of the time stamp and the date stamp by searching the analyzed area for pixel values that together form distinguishable text and have color values that substantially differ from surrounding pixel color values, and wherein the image editing unit is further configured to replace values of the undesirable pixels to remove the at least one of the time stamp and the date stamp.

19. A non-transitory computer readable storage medium on which is embedded one or more computer programs, said one or more computer programs implementing a method for processing an image containing undesirable pixels, said one or more computer programs comprising computer readable code for:

receiving a coarse identification of a location of the undesirable pixels, wherein the coarse identification includes identification of at least one undesirable pixel and at least one non-defective pixel in the image;

automatically determining an area in the image to be analyzed for undesirable pixel values according to the coarse identification received, wherein the area to be analyzed is further based upon a shape of the undesirable pixels being identified;

in the area determined to be analyzed, automatically classifying the pixels as one of undesirable and desirable; and

storing the classifications of the pixels.

20. The computer readable storage medium according to claim 19, said one or more computer programs further comprising computer readable code for:

replacing the pixels determined to be undesirable by replacing values of the undesirable pixels with replacement pixel values derived from a pixel value of at least one of the pixels determined to be desirable.