Edge based CMY automatic picture registration

• US 8,014,592 B2
• Filed: 06/27/2005
• Issued: 09/06/2011
• Est. Priority Date: 01/27/2005
• Status: Expired due to Fees

• Current Assignee: Thomson Licensing (Vantiva SA)
• Original Assignee: Thomson Licensing (Vantiva SA)
• Inventors: Lin, Shu
• Primary Examiner(s): Akhavannik; Hadi
• Application Number:
  US11/795,852
• Publication Number:
  US 20080118144A1
• Time in Patent Office: 2,262 Days
• Field of Search: 382/162, 382/199
• US Class Current: 382/162
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/33 using feature-based methods
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: H04N 1/3876 Recombination of partial images to recreate the original image
• 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 9/11 Scanning of colour motion picture films, e.g. for telecine

Abstract

An automatic process for performing CMY (Cyan, Magenta, Yellow) registration for film digital restoration. After three color components Cyan, Magenta, and Yellow of a picture are scanned into files, each component is divided into blocks, and edge detection is applied to each block, and an edge match is performed. The data of displacements is processed, and then affine transform parameters are calculated. The affine transform is then applied for each block, and warping is used to combine the color components and obtain the registered picture of a color component.

Claims

1. A method for automatically registering color components of a color film, the method comprising the steps of:

downsampling, from a desired resolution to a lower resolution, image data of the color components of the color film to be registered to speed up processing;

determining correlation data between the color components of the color film;

processing the correlation data;

upscaling matched displacement vectors to the desired resolution before combining is performed;

determining Affine Transform parameters for the color components;

calculating the Affine Transform for each pixel in the respective color component using the determined parameters; and

combining the color components to re-produce the color film.

2. The method according to claim 1, wherein said determining correlations comprises:

selecting a base color component;

calculating initial displacement coefficient vector values of the other color components with respect to said selected base color component;

dividing each picture frame into blocks;

detecting picture frame edges for the color components; and

matching the detected edges of each color component with respect to a chosen base color component.

3. The method according to claim 2, wherein said processing of the determined correlations comprises calculating new displacement vector values using the initially calculated displacement coefficient vector values.

4. The method according to claim 2, wherein said processing of the determined correlations comprises:

determining whether any large errors are present in the determined correlations;

modifying any large errors;

calculating new displacement value coefficients; and

re-calculating the displacement vector values using the newly calculated displacement value coefficients.

5. The method according to claim 4, wherein said determining is performed using a predetermined threshold.

6. The method according to claim 4, wherein said modifying errors is performed using by using neighboring block values or interpolation/extrapolation.

7. The method according to claim 4, wherein said calculating new displacement value coefficients is performed by applying a 3-order curve or 3-order plane to best fit numbers the error numbers in either direction.

8. The method according to claim 2, wherein said calculating the Affine transform comprises:

determining pixel position in the original picture for each pixel in a block using a nearest block analysis; and

defining affine transform parameters using the determined pixel positions and displacement values of the corresponding nearest blocks.

9. The method according to claim 1, wherein said combining comprises warping non base color components with the base color component to form a registered color image.

10. The method according to claim 9, further comprising converting the color components to a desired file or image format before forming the registered color image.