Apparatus and method for defect channel nulling

• US 6,442,301 B1
• Filed: 12/29/1997
• Issued: 08/27/2002
• Est. Priority Date: 01/06/1997
• Status: Expired due to Term

• Current Assignee: Kodak Alaris Inc (KPP (no. 2) Trustees Ltd.)
• Original Assignee: Applied Science Fiction, Inc.
• Inventors: Edgar, Albert D.
• Primary Examiner(s): LEE, TOMMY D
• Application Number:
  US08/999,421
• Time in Patent Office: 1,702 Days
• Field of Search: 382/254, 382/275, 358/463, 358/487, 358/506, 250/330, 250/341.1, 250/341.8, 250/339.11, 356/443, 356/444, 356/237.1, 356/239.1
• US Class Current: 382/275
• CPC Class Codes: H04N 23/81 for suppressing or minimising disturbance in the image signal generation
• CPC Class Codes: H04N 3/36 Scanning of motion picture films, e.g. for telecine
• CPC Class Codes: H04N 5/253 Picture signal generating by scanning motion picture films or slide opaques, e.g. for telecine scanning details therefor H04N3/36 ; standard conversion therefor H04N7/0112
• CPC Class Codes: H04N 9/11 Scanning of colour motion picture films, e.g. for telecine

Abstract

A method of and apparatus for removing the effects of surface and near surface image storage media defects from a scanned image using an infrared record as a norming control. Each pixel in a visible channel of the scanned images is divided by the corresponding pixel in the associated infrared control channel after it has been altered in gain to match the degree of defect in the visible channel. By appropriately altering the gain prior to dividing the pixel information, imbalances between the visible and infrared records which would leave defect residue after the division are reduced or eliminated. To remove defect residue, a degree of nulling is established for each defect region based on the visible and infrared content in that region. In one embodiment, the articulation gain is multiplicatively applied to the logarithm of the visible and infrared records. To prevent variations in articulation gain from mottling brightness, the articulation gain is applied to details extracted through a high pass filter. In this embodiment, the image is divided into frequency bands in a pyramid structure, and separate articulation gain is applied to each band. The articulation gains are calculated using the crosscorrelation between the visible and infrared images within a region to null the component of the visible channel matching the defect detected by the infrared channel. Gain limits are set based on correlations measured over the entire image.

Claims

1. A method for correcting defects in an image comprising:

applying light to an image storing medium;

receiving from the image storing medium an image signal indicating an image deteriorated by defects;

applying a defect distinguishing radiation to the image storing medium;

receiving from the image storing medium a defect signal, in register with the image signal, indicating the defects in the image;

applying a gain to the defect signal to generate a normalized defect signal;

differencing the image signal and normalized defect signal to generate a recovered image signal; and

adjusting the gain to minimize the defects in the recovered image signal.

2. The method of claim 1 wherein the image storing medium is a substrate bearing a two dimensional image.

3. The method of claim 2 wherein the substrate bears a visible image.

4. The method of claim 2 wherein the defect signal indicates the locations and magnitudes of substrate defects.

5. The method of claim 2 wherein the defect signal is received from the substrate in infrared light.

6. The method of claim 1 wherein the step of adjusting the gain to minimize the defects in the recovered image signal comprises the step of measuring the defects in the recovered image signal.

7. The method of claim 6 wherein the step of measuring the defects in the recovered image signal comprises the step of calculating the correlation between the recovered image signal and the defect signal.

8. The method of claim 7 wherein the step of adjusting the gain further comprises the step of altering the gain as a function of the correlation between the recovered image signal and the defect signal.

9. The method of claim 1 wherein the defect signal includes a plurality of defect signal frequency bands, and wherein the gain is adjusted differently between at least two of the bands.

10. The method of claim 9 wherein one of the defect signal frequency bands comprises a defect record of the image signal.

11. The method of claim 9 wherein the defect signal frequency bands are logarithmically spaced by a frequency incrementing factor.

12. The method of claim 11 wherein the frequency incrementing factor is two.

13. The method of claim 9 wherein the defect signal frequency bands are linearly spaced.

14. The method of claim 9 wherein the step of adjusting the gain to minimize the defects in the recovered image signal comprises the step of calculating the correlation between the image signal and at least one of the defect signal frequency bands.

15. The method of claim 14 wherein the step of adjusting the gain further comprises the step of calculating the gain as a function of the correlation between the image signal and at least one of the defect signal frequency bands.

16. The method of claim 9 wherein the defect signal received is further characterized to include a plurality of defect signal regions, and wherein the gain is adjusted differently between at least two of the defect signal regions.

17. The method of claim 16 wherein the step of adjusting the gain to minimize the defects in the recovered image signal comprises the step of calculating the correlation between the image signal and at least two of the defect signal regions.

18. The method of claim 17 wherein the step of adjusting the gain comprises the step of calculating the gain as a function of the correlation between the image signal and at least two of the defect signal regions.

19. The method of claim 18 wherein the step of calculating the gain for a selected defect signal region within a selected defect signal frequency band comprises the steps of calculating the autocorrelation of the defect signal within the selected defect signal region and the defect signal band, and the crosscorrelation of the defect signal within the selected defect signal region and the defect signal band with the image signal.

20. The method of claim 19 wherein the gain within the selected defect signal region and the defect signal band is a function of said crosscorrelation divided by said autocorrelation.

21. The method of claim 20 wherein the defect signal regions are further characterized as comprising of a plurality of points, and wherein the steps of calculating the autocorrelation and the crosscorrelation further includes the steps of calculating the point autocorrelations of the defect signal at the plurality of points and the point cross correlations between the defect signal and the registered image signal at the plurality of points, and averaging the point autocorrelations to obtain the autocorrelation within a region and averaging the point cross correlations to obtain the cross correlations within a region.

22. The method of claim 21 wherein the averaging steps further include spatially lowpass filtering the point autocorrelations to produce a smudged autocorrelation signal containing smudged autocorrelation points, and spatially lowpass filtering the point cross correlations to produce a smudged crosscorrelation signal containing smudged crosscorrelation points.

23. The method of claim 22 wherein the adjusted gain at a selected point of the defect signal is a function of the corresponding smudged autocorrelation point divided by the corresponding smudged crosscorrelation point.

24. The method of claim 16 wherein the defect signal received is further characterized as comprising at least one intermediate region overlapping at least two of the defect signal regions, and wherein the gain is further characterized as changing smoothly across at least two of the defect signal regions and the intermediate region.

25. An apparatus for correcting the effect of defects in an image comprising:

means for applying light to the image storage medium;

means for receiving from the image storing medium an image signal indicating an image deteriorated by defects in the image storage medium;

means for applying a defect distinguishing radiation to the image storage medium;

means for receiving from the image storage medium a defect signal, in register with the image signal, indicating the defects in the image;

means for applying a gain to the defect signal to generate a normalized defect signal;

means for differencing the image signal and normalized defect signal to generate a recovered image signal; and

means for adjusting the gain to minimize the defects in the recovered image signal.

26. The apparatus of claim 25 wherein the recording medium is a substrate bearing a two dimensional image.

27. The apparatus of claim 26 wherein the substrate contains a visible image.

28. The apparatus of claim 26 wherein the defect signal is received from the substrate in infrared.

29. The apparatus of claim 25 wherein the gain adjusting means includes means for measuring the defects in the recovered image signal.

30. The apparatus of claim 29 wherein the measuring means comprises means for calculating the correlation between the recovered image signal and the defect signal.

31. The apparatus of claim 25 wherein the gain adjusting means comprises means for altering the gain as a function of the correlation between the recovered image signal and the defect signal.

32. The apparatus of claim 25 wherein the gain adjusting means comprises means for measuring the defects in the image signal.

33. The apparatus of claim 32 wherein the measuring means comprises means for calculating the correlation between the image signal and the defect signal.

34. The apparatus of claim 33 wherein the gain adjusting means comprises means for calculating the gain as a function of the correlation between the image signal and the defect signal.

35. The apparatus of claim 25 wherein the defect signal comprises a plurality of defect signal frequency bands, and wherein the gain is adjusted differently between at least two of the bands.

36. The apparatus of claim 35 wherein one of the defect signal frequency bands is a mean of the defect signal.

37. The apparatus of claim 35 wherein the defect signal frequency bands are logarithmically spaced by a frequency incrementing factor.

38. The apparatus of claim 37 wherein the frequency incrementing factor is two.

39. The apparatus of claim 35 wherein the defect signal frequency bands are linearly spaced.

40. The apparatus of claim 35 wherein the gain adjusting means further comprises means for calculating the correlation between the image signal and at least one of the defect signal frequency bands.

41. The apparatus of claim 40 wherein the gain adjusting means further comprises means for calculating the gain as a function of the correlation between the image signal and at least one of the defect signal frequency bands.

42. The apparatus of claim 35 wherein the defect signal is further characterized by comprising a plurality of defect signal regions, and wherein the gain is adjusted differently between at least two of the regions.

43. A method of processing visual image data with non-image defect data to remove non-image defect information from the visual image data, the non-image defect data being registered with the visual image data, comprising:

creating, from the visual image data, a plurality of visual image data sets for respective spatial frequency bands;

creating from the visual image data a plurality of non-image defect data sets for respective spatial frequency bands substantially similar to the spatial frequency bands for the visual imaging data;

cross correlating the visual image data set and the non-image defect data sets for respective spatial frequency bands to obtain crosscorrelation variables for each of the spatial frequency bands;

autocorrelating the non-image defect data set for each of the spatial frequency bands to obtain autocorrelation variables for each of the spatial frequency bands;

processing the crosscorrelation variables over a plurality of spatial regions to obtain spatially defined crosscorrelation variables for each of the spatial frequency bands;

processing the autocorrelation variables over a plurality of spatial regions substantially similar to the spatial regions of the crosscorrelation variables to obtain spatially defined autocorrelation variables for each of the spatial frequency bands;

processing the spatially defined crosscorrelation variables with the spatially defined autocorrelation variables to obtain spatially defined nulling gains for each of the spatial frequency bands;

applying the spatially defined nulling gains to the respective non-image defect data sets to obtain nulling data sets for the respective spatial frequency bands;

applying the nulling data sets to the visual image data sets for the respective spatial frequency bands to obtain corrected visual image data sets for the respective spatial frequency bands; and

merging the corrected visual image data sets for the respective spatial frequency bands to obtain corrected visual image data.

44. The method of claim 43 further comprising identifying a plurality of spatial frequency bands to overcome limitations in real optics and image sensors, the visual image data further comprising a plurality of respective visual image data sets for the spatial frequency bands, the non-image defect data further comprising a plurality of respective non-image defect data sets for the spatial frequency bands, the nulling gain further comprising a plurality of respective nulling gain sets for the spatial frequency bands, and the nulling data further comprising a plurality of respective nulling data sets for the spatial frequency bands.