INTERPOLATION SYSTEM AND METHOD

• US 20090252411A1
• Filed: 09/09/2008
• Published: 10/08/2009
• Est. Priority Date: 04/08/2008
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Siddiqui, Hasib Ahmed, Hwang, Hau
• Granted Patent: US 8,229,212 B2
• Time in Patent Office: Days
• US Class Current: 382/167
• CPC Class Codes: G06T 3/4015 Image demosaicing, e.g. colour filter arrays [CFA] or Bayer patterns

Abstract

Interpolations systems and methods are disclosed. In a particular embodiment, a system is disclosed that includes an input to receive image data. The system also includes an image processing system responsive to the image data and including a demosaicing module. The demosaicing module is configured to use adaptive bi-cubic spline interpolation. The system further includes an output responsive to the image processing system and adapted to provide output data.

Claims

1. A system comprising:

an input to receive image data;

an image processing system responsive to the image data and including a demosaicing module, the demosaicing module configured to use adaptive bi-cubic spline interpolation; and

an output responsive to the image processing system and adapted to provide output data.

2. The system of claim 1, wherein the demosaicing module is configured to receive input image data including first color data representing a first color, second color data representing a second color, and third color data representing a third color, and wherein the demosaicing module includes an interpolation filter configured to receive the input image data and to output filtered data including interpolated first color data, wherein the interpolation filter is adapted to perform adaptive bicubic spline interpolation of fourth color data representing a fourth color.

3. The system of claim 2, wherein the demosaicing module further includes:

a feature extraction module to process the input image data and to generate feature data corresponding to an image feature; and

a classifier module responsive to the feature data to provide a classifier output associated with an orientation of the image feature.

4. The system of claim 3, wherein the interpolation filter includes:

logic to interpolate the second color data to generate interpolated second color data;

logic to subtract the interpolated second color data from the first color data to generate the fourth color data;

adaptive bi-cubic spline interpolation logic configured to receive the fourth color data and to generate interpolated fourth color data; and

logic to add the second color data to the interpolated fourth color data to produce interpolated first color data at locations associated with the second color data.

5. The system of claim 4, wherein the first color is green, wherein the second color is red or blue, and wherein the fourth color is yellow or cyan.

6. The system of claim 4, wherein the first color is yellow, wherein the second color is magenta or cyan, and wherein the fourth color is red or blue.

7. The system of claim 4, wherein the feature extraction module is configured to perform derivative operations associated with the input image data, and wherein the feature data includes three dimensional data having values of three directional derivative operations or including twenty-seven dimensional data having values of twenty-seven directional derivative operations.

8. The system of claim 7, wherein the classifier module is configured to selectively determine the classifier output using a Gaussian mixture model type algorithm when the feature data includes the three dimensional data or using an Adaboost type algorithm when the feature data includes the twenty-seven dimensional data.

9. The system of claim 4, wherein the adaptive bi-cubic spline interpolation logic is configured to generate an initial value of the interpolated fourth color data using the classifier output and to interpolate the fourth color data using the initial value and an adaptive weighting function.

10. The system of claim 4, wherein the interpolation filter includes:

logic to interpolate the third color data to generate interpolated third color data;

logic to subtract the interpolated third color data from the first color data to generate fifth color data,

wherein the adaptive bi-cubic spline interpolation logic is further configured to receive the fifth color data and to generate interpolated fifth color data; and

logic to add the third color data to the interpolated fifth color data to produce interpolated first color data at locations associated with the third color data.

11. The system of claim 10, wherein the first color is green, wherein the second color is red, wherein the third color is blue, wherein the fourth color is yellow, and wherein the fifth color is cyan.

12. The system of claim 10, wherein the first color is yellow, wherein the second color is magenta, wherein the third color is cyan, wherein the fourth color is red, and wherein the fifth color is blue.

13. The system of claim 10, wherein the adaptive bi-cubic spline interpolation logic is configured to generate an initial value of the interpolated fourth color data using the classifier output and to interpolate the fourth color data using the initial value of the interpolated fourth color data and an adaptive weighting function, and wherein the adaptive bi-cubic spline interpolation logic is configured to generate an initial value of the interpolated fifth color data using the classifier output and to interpolate the fifth color data using the initial value of the interpolated fifth color data and the adaptive weighting function.

14. The system of claim 1, wherein the image processing system includes:

a gamma module to generate gamma corrected data from data received from the demosaicing module;

a color calibration module to perform a calibration on the gamma corrected data;

a color space conversion module coupled to convert an output of the color calibration module to a color space; and

a compress and store module to receive an output of the color space conversion module and to store compressed output data to the image storage device.

15. An interpolation system comprising:

an interpolation filter configured to receive first color data and including adaptive bi-cubic spline interpolation logic configured to perform an interpolation operation using a rotated weighting grid; and

wherein the first color data represents a first color of input image data, the input image data further including second color data representing a second color and third color data representing a third color, the first color appearing more frequently in the input image data than the second color and the third color.

16. The interpolation system of claim 15, further comprising:

a feature extraction module configured to receive the input image data and to output feature data associated with directional derivatives of the input image data; and

a classifier module configured to receive the feature data and to generate a classifier output associated with an orientation of an image feature associated with the input image data;

wherein the interpolation filter is configured to generate interpolated first color data using the classifier output.

17. The interpolation system of claim 16, wherein the adaptive bi-cubic spline interpolation logic includes an initial value estimator module configured to estimate an initial value at a particular location of the input image data by applying the classifier output to values at locations neighboring the particular location.

18. The interpolation system of claim 17, wherein the adaptive bi-cubic spline interpolation logic further includes an adaptive weighting function module to generate an interpolated value at the particular location using a discounted value of a nearby location by applying a weighting function to a difference between the value of the nearby location and the initial value.

19. The interpolation system of claim 18, wherein the weighting function includes an exponentiation to a power of four of the difference between the value of the nearby location and the initial value.

20. The interpolation system of claim 15, wherein the adaptive bi-cubic spline interpolation logic is configured to generate an interpolated value at a particular location using weighted values of sixteen nearby locations by applying the rotated weighting grid to values at the sixteen nearby locations, the rotated weighting grid indicating one of a square of a first weighting factor, a square of a second weighting factor, and a product of the first weighting factor and the second weighting factor.

21. The interpolation system of claim 15, wherein the interpolation filter further includes:

fourth color data production logic configured to generate fourth color data by subtracting second interpolated color data from the first color data, the fourth color production logic configured to provide the fourth color data to the adaptive bi-cubic spline interpolation logic to produce interpolated fourth color data; and

interpolated first color data production logic configured to receive the interpolated fourth color data from the adaptive bi-cubic spline interpolation logic and to add the second color data to generate interpolated first color data.

22. The interpolation system of claim 21, wherein the interpolation filter further includes fifth color data production logic configured to generate fifth color data by subtracting third interpolated color data from the first color data, the fifth color production logic configured to provide the fifth color data to the adaptive bi-cubic spline interpolation logic to produce interpolated fifth color data, and wherein the interpolated first color data production logic is further configured to receive the interpolated fifth color data from the adaptive bi-cubic spline interpolation logic and to add the third color data to generate interpolated first color data at locations corresponding to the third color.

23. The interpolation system of claim 15, wherein the input image data includes Bayer mosaic data, and further comprising:

a first difference channel coupled to receive interpolated first color data from the interpolation filter and configured to provide first difference channel output data corresponding to the second color; and

a second difference channel coupled to receive the interpolated first color data from the interpolation filter and configured to provide second difference channel output data corresponding to the third color,

wherein the interpolated first color data, the first difference channel output data, and the second difference channel output data provide a result of a demosaicing operation on the input image data.

24. A method comprising:

receiving image data including first color data representing a first color, second color data representing a second color, and third color data representing a third color, the image data including more of the first color data than the second color data and the third color data;

generating fourth color data that represents a fourth color by subtracting interpolated second color data from the first color data;

interpolating the fourth color data to generate interpolated fourth color data; and

generating interpolated first color data using the interpolated fourth color data added to the second color data.

25. The method of claim 24, wherein the fourth color is yellow, cyan, red, or blue.

26. The method of claim 24, further comprising:

generating fifth color data that represents a fifth color by subtracting interpolated fifth color data from the first color data;

interpolating the fifth color data to generate interpolated fifth color data; and

generating interpolated first color data using the interpolated fifth color data added to the third color data.

27. The method of claim 24, wherein interpolating the fourth color data includes performing an adaptive bi-cubic spline interpolation operation using sixteen nearest neighbors.

28. The method of claim 27, wherein the adaptive bi-cubic spline interpolation operation adjusts an interpolation value based on an exponentiation to a power of four of a difference of an initial value and a nearby value.

29. A processor readable medium storing processor instructions that are executable by the processor to:

receive image data including first color data representing a first color, second color data representing a second color, and third color data representing a third color;

generate interpolated second color data;

subtract the interpolated second color data from the first color data to generate fourth color data representing a fourth color;

interpolate the fourth color data to generate interpolated fourth color data using an adaptive bi-cubic spline interpolation operation;

add the second color data to the interpolated fourth color data to produce interpolated first color data; and

store the interpolated first color data to a memory.

30. The processor readably medium of claim 29, wherein the processor instructions further include instructions executable by the processor to:

generate interpolated yellow data using the adaptive bi-cubic spline interpolation operation when the first color is green and the second color is red;

generate interpolated cyan data using the adaptive bi-cubic spline interpolation operation when the first color is green and the second color is blue;

generate interpolated red data using the adaptive bi-cubic spline interpolation operation when the first color is yellow and the second color is magenta; and

generate interpolated blue data using the adaptive bi-cubic spline interpolation operation when the first color is yellow and the second color is cyan.

31. The processor readable medium of claim 30, wherein the adaptive bi-cubic spline interpolation operation applies a rotated weighting grid to adjust values of the fourth color data at predetermined positions relative to a particular location.

32. The processor readable medium of claim 31, wherein the adaptive bi-cubic spline interpolation operation divides each value of the fourth color at the predetermined positions by a difference between an initial value of the particular location and the value of the fourth color at the predetermined position, the difference raised to the fourth power.