Detecting redeye defects in digital images

• US 8,036,458 B2
• Filed: 11/08/2007
• Issued: 10/11/2011
• Est. Priority Date: 11/08/2007
• Status: Active Grant

• Current Assignee: FotoNation Limited (Adeia Inc.)
• Original Assignee: DigitalOptics Corporation Europe Limited (Adeia Inc.)
• Inventors: Ciuc, Mihai
• Primary Examiner(s): Ge; Yuzhen
• Application Number:
  US11/937,377
• Publication Number:
  US 20090123063A1
• Time in Patent Office: 1,433 Days
• Field of Search: None
• US Class Current: 382/167
• CPC Class Codes: G06V 40/193 Preprocessing; Feature extraction

Abstract

A method for detecting a redeye defect in a digital image containing an eye comprises converting the digital image into an intensity image, and segmenting the intensity image into segments each having a local intensity maximum. Separately, the original digital image is thresholded to identify regions of relatively high intensity and a size falling within a predetermined range. Of these, a region is selected having substantially the highest average intensity, and those segments from the segmentation of the intensity image whose maxima are located in the selected region are identified.

Claims

1. A method for detecting a redeye defect in a digital image containing an eye, comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.

2. The method of claim 1 further comprising identifying at least one of the one or more regions having a size falling within a predetermined range, and wherein said region of substantially the highest average intensity is selected from said at least one of the one or more regions.

3. The method of claim 1 wherein the identifying comprises identifying one or more of said segments whose maxima are located in said region.

4. The method of claim 1, comprising subjecting the intensity image to local averaging prior to mean shift segmentation.

5. The method of claim 1, wherein the converting comprises mean shift segmentation of the intensity image of the digital image.

6. The method of claim 1, comprising subjecting the intensity image to local averaging prior to mean shift segmentation.

7. The method of claim 1, further comprising:

computing an intensity histogram of a portion of the intensity image corresponding to the one or more of said segments,

smoothing said histogram to generate a smoothed histogram,

determining a threshold from the smoothed histogram, and

thresholding the intensity histogram with the threshold from the smoothed histogram.

8. The method of claim 1, further comprising applying the converting and segmenting to one or more portions of the image suspected to contain an eye.

9. A digital image acquisition and processing apparatus including means for performing the method claimed in claim 1.

10. A method for detecting a redeye defect in a digital image containing an eye, comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion,

wherein the average intensities of two regions of the one or more regions having the

highest and next highest average intensities differ by less than a predetermined amount, and said region comprises the region having the highest average saturation.

11. A method for detecting a redeye defect in a digital image containing an eye, comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion,

wherein the thresholding uses a threshold obtained by identifying a highest intensity pixel

inside the intensity image, computing an average intensity of the pixels inside a smaller rectangle

centered on the highest intensity pixel, and assigning a threshold to a sum of the average

intensity and a fixed quantity.

12. The method of claim 11, wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.

13. One or more non-transitory processor readable media having code embedded therein for programming a processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.

14. The one or more processor readable media of claim 13 wherein the method further comprises identifying at least one of the one or more regions having a size falling within a predetermined range, and wherein said region of substantially the highest average intensity is selected from said at least one of the one or more regions.

15. The one or more processor readable media of claim 13 wherein the identifying comprises identifying one or more of said segments whose maxima are located in said region.

16. The one or more processor readable media of claim 13, wherein the method further comprises subjecting the intensity image to local averaging prior to mean shift segmentation.

17. The one or more processor readable media of claim 13, wherein the converting comprises mean shift segmentation of the intensity image of the digital image.

18. The one or more processor readable media of claim 13, wherein the method further comprises subjecting the intensity image to local averaging prior to mean shift segmentation.

19. The one or more processor readable media of claim 13, wherein the method further comprises:

computing an intensity histogram of a portion of the intensity image corresponding to the one or more of said segments,

smoothing said histogram to generate a smoothed histogram,

determining a threshold from the smoothed histogram, and

thresholding the intensity histogram with the threshold from the smoothed histogram.

20. The one or more processor readable media of claim 13, wherein the method further comprises applying the converting and segmenting to one or more portions of the image suspected to contain an eye.

21. One or more non-transitory processor readable media having code embedded therein for programming a processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by less than a predetermined amount, and said region comprises the region having the highest average saturation.

22. One or more non-transitory processor readable media having code embedded therein for programming a processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising:

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having the highest average intensity, and

identifying one or more of said segments intersecting the region according to a predetermined criterion,

wherein the thresholding uses a threshold obtained by identifying a highest intensity pixel inside the intensity image, computing an average intensity of the pixels inside a smaller rectangle centered on the highest intensity pixel, and assigning a threshold to a sum of the average intensity and a fixed quantity.

23. The one or more processor readable media of claim 22, wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.

24. A digital image acquisition device, comprising:

a lens and image sensor for acquiring digital images;

a processor; and

a memory having processor readable code embedded therein for programming the processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising;

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having substantially the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.

25. The device of claim 24 wherein the method further comprises identifying at least one of the one or more regions having a size falling within a predetermined range, and wherein said region of substantially the highest average intensity is selected from said at least one of the one or more regions.

26. The device of claim 24 wherein the identifying comprises identifying one or more of said segments whose maxima are located in said region.

27. The device of claim 24, wherein the method further comprises subjecting the intensity image to local averaging prior to mean shift segmentation.

28. The device of claim 24, wherein the converting comprises mean shift segmentation of the intensity image of the digital image.

29. The device of claim 24, wherein the method further comprises subjecting the intensity image to local averaging prior to mean shift segmentation.

30. The device of claim 24, wherein the method further comprises:

computing an intensity histogram of a portion of the intensity image corresponding to the one or more of said segments,

smoothing said histogram to generate a smoothed histogram,

determining a threshold from the smoothed histogram, and

thresholding the intensity histogram with the threshold from the smoothed histogram.

31. The device of claim 24, wherein the method further comprises applying the converting and segmenting to one or more portions of the image suspected to contain an eye.

32. A digital image acquisition device, comprising:

a lens and image sensor for acquiring digital images;

a processor; and

a memory having processor readable code embedded therein for programming the processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising;

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having substantially the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by less than a predetermined amount, and said region comprises the region having the highest average saturation.

33. A digital image acquisition device, comprising:

a lens and image sensor for acquiring digital images;

a processor; and

a memory having processor readable code embedded therein for programming the processor to perform a method for detecting a redeye defect in a digital image containing an eye, the method comprising;

converting the digital image into an intensity image,

segmenting at least a portion of the intensity image into segments each having a local intensity maximum,

thresholding a corresponding portion of the digital image to identify one or more regions of relatively high intensity,

selecting a region from the one or more regions having substantially the highest average intensity, and

identifying one or more of said segments intersecting the region according to a pre-determined criterion, and

wherein the thresholding uses a threshold obtained by identifying a highest intensity pixel inside the intensity image, computing an average intensity of the pixels inside a smaller rectangle centered on the highest intensity pixel, and assigning a threshold to a sum of the average intensity and a fixed quantity.

34. The device of claim 33, wherein the average intensities of two regions of the one or more regions having the highest and next highest average intensities differ by more than a predetermined amount, and said region comprises the region having the highest average intensity.