Automatic cross color elimination

• US 5,509,086 A
• Filed: 12/23/1993
• Issued: 04/16/1996
• Est. Priority Date: 12/23/1993
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Edgar, Albert D., Kasson, James M.
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Tran, Phuoc
• Application Number:
  US08/173,798
• Time in Patent Office: 845 Days
• Field of Search: 382/1, 382/17, 382/100, 382/162, 382/167, 382/191, 358/518, 358/515, 358/527, 355/35, 355/77, 348/64, 348/263
• US Class Current: 382/167
• CPC Class Codes: H04N 1/58 Edge or detail enhancement; Noise or error suppression, e.g. colour misregistration correction H04N1/62 takes precedence
• CPC Class Codes: H04N 1/6011 with simulation on a subsidiary picture reproducer H04N1/622 takes precedence; matching two or more picture reproducers H04N1/6052

Abstract

Color crosstalk is determined between layers of an image storage medium based on the cross correlations and autocorrelations of noise in the grain pattern in each layer of the image storage medium. Rather than relying on prior measurement under laboratory conditions, the invention scans the storage medium in a plurality of spectral bands to derive the record stored in each of the dye layers. A computer derives the autocorrelations and the crosscorrelations of a group of pixels between the plurality of color records of the scanned images each of which corresponds to one of the spectral bands. The invention is based on the observation that as each dye layer is deposited separately in the film, one would expect that the "noise" from the grain boundaries in one layer to be independent from the "noise" due to the grain boundaries in other layers. If there is correlation in noise between separate color scans, it is because the scanner is measuring the grain boundaries in more than one layer in the color scans.

Claims

1. A method for correcting for the effects of crosscolor crosstalk between colors of a color image stored in an image storage medium and an imaging device used to scan the color image into a plurality of scanned images each corresponding to a color of the color image, comprising the steps of:

identifying noise patterns in the plurality of scanned images;

calculating crosscolor correlations between the identified noise patterns in the plurality of scanned images; and

correcting for color crosstalk in each of the scanned images according to the crosscolor correlations.

2. The method as recited in claim 1 wherein the colors comprise ranges for red, blue and green spectra.

3. The method as recited in claim 2 wherein the colors further comprise a range for infrared spectra.

4. The method as recited in claim 1 which further comprises the steps of:

determining a set of autocorrelations between the image and the imaging device in each of the scanned images; and

correcting each of the scanned images according to the autocorrelations.

5. The method as recited in claim 1 wherein each of the scanned images corresponds to a layer in a set of layers in the image storage medium.

6. A method for reducing the effects of color crosstalk between a color image and a scanner used to scan the color image, comprising the steps of:

successively scanning the color image with the scanner to produce a set of scanned images of the color image each of which corresponds to a different color;

separating noise patterns from image patterns in the set of scanned images;

calculating crosscolor correlations between the separated noise patterns in pairs of the set of scanned images;

calculating a set of correction coefficients according to the calculated crosscolor correlations; and

using the set of correction coefficients to produce a set of adjusted images corrected for color crosstalk from the set of scanned images.

7. The method as recited in claim 6 wherein three of the set of scanned images respectively correspond to a red, a green and a blue scan of the color image, of a cyan, a magenta and a yellow layer in color substrate.

8. The method as recited in claim 7 which further comprises the steps of:

identifying selected areas of the color image which have relatively little signal component other than noise; and

calculating crosscolor correlations between noise patterns in the selected areas.

9. The method as recited in claim 7 which further comprises the step of correcting a scanned image according to the set of correction coefficients.

10. The method as recited in claim 7 which further comprises the steps of:

processing color values from the scanned images in an iterative loop which develops a succession of color estimates to determine all amount of each of the different colors in the color image;

processing color values from the scanned images in an iterative loop which develops a succession of crosscorrelation coefficients; and

generating a spectral response characterization of the image storage medium according to the crosscorrelation coefficients.

11. The method as recited in claim 6 which further comprises the steps of:

processing color values from the scanned images in an iterative loop which develops a succession of color estimates to determine an amount of each of the different colors in the color image;

processing color values from the scanned images in an iterative loop which develops a succession of crosscorrelation coefficients; and

generating a spectral response characterization of the scanner according to the crosscorrelation coefficients.

12. A method for correcting for the effects of crosstalk between images in register with each other, each registered image consisting of a desired pattern and a noise pattern, comprising the steps of;

introducing crosstalk between the registered images to produce adjusted images;

substantial separating a portion of a noise pattern from each adjusted image;

calculating crosscorrelations of the separated noise patterns of the adjusted images; and

adjusting the introduced crosstalk so that crosscorrelations between noise patterns of the adjusted images tend toward zero.

13. The method as recited in claim 12 wherein the registered images include the respective red, green, and blue images of a colored image.

14. The method as recited in claim 13 further including an infrared image.

15. The method as recited in claim 12 wherein the separation step further comprises the step of applying a spatial high frequency filter to the image.

16. The method as in claim 15 wherein the separation step further comprises the step of suppressing the desired pattern from inclusion in the noise pattern in regions of the registered images in which the desired pattern is stronger than in other regions of the registered images.

17. The method as recited in claim 16 wherein a strength of the desired pattern in a region is measured as a function of the magnitude of the image remaining in the region after applying a spatial high frequency filter.

18. The method as recited in claim 12 wherein the introducing step further comprised the steps of;

changing the intensity of at least two registered images by selected magnitudes, which may be positive or negative, to produce at least two intensified images; and

adding the at least two intensified images together.

19. The method as recited in claim 18 wherein the registered images are first acted on by a function which expands a grayscale range by translating a darkest pixel to pure black and a lightest pixel to pure white to produce a new contrast translation curve before changing pixel intensity.

20. The method of claim 19 wherein the adjusted image is acted on by an inverse function to the function which expands a grayscale range applied to the registered images.

21. The method as recited in claim 18 wherein the selected magnitudes are chosen by an iterative process so that the crosscorrelations tend to zero.

22. The method as recited in claim 12 wherein the registered images are derived from a colored film and wherein the noise pattern is the grain pattern of the colored film.

23. The method as recited in claim 12 wherein the registered images are placed in a computer memory by a scanning process.

24. A system for minimizing crosstalk while scanning at least three images in register with each other on a film substrate, each image consisting of a desired pattern and a grain pattern, comprising:

means for scanning the at least three images into computer memory to produce scanned images, the scanning means introducing undesired crosstalk between the images as they exist on the film substrate into the scanned images;

means for substantially isolating a portion of the grain pattern from the desired pattern in each scanned image, to produce an isolated portion of the grain pattern for each scanned image; and

means for adjusting the amount of crosstalk between the scanned images to produce adjusted images such that a proportional adjustment of crosstalk between the isolated portions of the grain patterns would cause the crosscorrelations between the isolated portions of the grain patterns to tend toward zero.

25. A system for correcting for the effects of crosstalk between images in register with each other, each registered image consisting of a respective desired pattern and a respective noise pattern, comprising:

means for introducing crosstalk between the registered images to produce adjusted images;

means for substantially separating a portion of a noise pattern from the adjusted images;

means for calculating crosscorrelations between the separated noise patterns of the adjusted images; and

means for adjusting the introduced crosstalk so that crosscorrelations between the separated noise patterns of the adjusted images tend toward zero.

26. A system for correcting for the effects of crosscolor crosstalk between colors of a color image stored in an image storage medium and an imaging device used to scan the color image into a plurality of scanned images each corresponding to a color of the color image, comprising

means for identifying noise patterns in the plurality of scanned images;

means for calculating crosscorrelations between the identified noise patterns in pairs of the scanned images; and

means for correcting for color crosstalk in each of the scanned images according to the crosscolor correlations between noise patterns.