Image distortion correction

• US 9,177,368 B2
• Filed: 12/17/2007
• Issued: 11/03/2015
• Est. Priority Date: 12/17/2007
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Cabral, Brian K., Lin, Shang-Hung, Tjandrasuwita, Ignatius
• Primary Examiner(s): ENTEZARI, MICHELLE M
• Application Number:
  US12/002,562
• Publication Number:
  US 20090154822A1
• Time in Patent Office: 2,878 Days
• Field of Search: 382/255
• US Class Current: 1/1
• CPC Class Codes: G06T 5/80 Geometric correction

Abstract

Methods and systems for reducing or eliminating distortion in an image are described. The approach generally involves determining the distortion introduced by a lens, and modifying a captured image to reduce that distortion. In one embodiment, the distortion information associated with a lens is determined. The distortion information is stored. A captured image taken by that lens is processed, with reference to the distortion information.

Claims

1. A method of reducing image distortion, said method comprising:

determining distortion information associated with a lens system, wherein said distortion information is associated with distortion at a plurality of focal lengths, and wherein said distortion information is further associated with color distortion;

storing said distortion information;

processing a captured image taken with said lens system, wherein said processing is performed with reference to said distortion information; and

generating a corrected image comprising reduced image distortion compared to said captured image, wherein a shader of a graphics processing unit (GPU) is configured to perform said generating by calculating a corrected pixel value for each pixel in said corrected image based on a correction value for each pixel, wherein said correction value is obtained from a calibration texture, wherein said calibration texture is operable to correct a distortion introduced by said lens system, and wherein said shader is operable to be programmed by a texturing engine to use said calibration texture to output each pixel in a proper location in said corrected image.

2. The method of claim 1, wherein said determining comprises:

taking a calibration picture of a reference image using said lens system; and

calculating said distortion information by comparing said calibration picture with said reference image.

3. The method of claim 2 further comprising:

repeating said taking and said calculating for a plurality of reference images, to determine said distortion information for a plurality of distortion types.

4. The method of claim 1, wherein said generating comprises:

extracting a correction value for a destination pixel in said corrected image from said distortion information;

obtaining a source pixel value from said captured image;

modifying said source pixel value with reference to said correction value to generate said destination pixel; and

repeating said extracting, said obtaining, and said modifying for each of a plurality of destination pixels in said corrected image.

5. The method of claim 1, wherein said storing comprises:

writing said distortion information as a correction image.

6. The method of claim 5, wherein said correction image comprises a plurality of correctional pixel data.

7. The method of claim 6, wherein said correctional pixel data comprises position translation information.

8. The method of claim 7, wherein said processing comprises:

obtaining a source pixel value from said captured image;

translating, with reference to said position translation information, said source pixel value from a first location in said captured image to a second location in a corrected image; and

repeating said obtaining and said translating for a plurality of destination pixels in said corrected image.

22. The method of claim 1, wherein said distortion at a plurality of focal lengths and said distortion associated with color distortion are corrected simultaneously, wherein at least two different correction processes are carried out, and said at least two different correction processes are carried out simultaneously.

24. The method of claim 1 wherein said correction value further comprises information regarding adjusting a location of a pixel.

9. A computer-readable storage medium having computer-readable program code embodied therein for causing a computer system to perform a method of reducing image distortion, said method comprising:

determining distortion information associated with a lens system, wherein said distortion information is associated with distortion at a plurality of focal lengths, and wherein said distortion information is associated with color distortion;

storing said distortion information;

processing a captured image taken with said lens system, wherein said processing is performed with reference to said distortion information; and

generating a corrected image comprising reduced image distortion compared to said captured image, wherein a shader of a graphics processing unit (GPU) is configured to perform said generating by calculating a corrected pixel value for each pixel in said corrected image based on a correction value for each pixel, wherein said correction value is obtained from a calibration texture, wherein said calibration texture is operable to correct a distortion introduced by said lens system, and wherein said shader is operable to be programmed by a texturing engine to use said calibration texture to output each pixel in a proper location in said corrected image.

10. The computer-readable storage medium of claim 9, wherein said determining comprises:

taking a calibration picture of a reference image using said lens system; and

calculating said distortion information by comparing said calibration picture with said reference image.

11. The computer-readable storage medium of claim 10, wherein said method further comprises:

repeating said taking and said calculating for a plurality of reference images, to determine said distortion information for a plurality of distortion types.

12. The computer-readable storage medium of claim 9, wherein said generating comprises:

extracting a correction value for a destination pixel in said corrected image from said distortion information;

obtaining a source pixel value from said captured image;

modifying said source pixel value with reference to said correction value to generate said destination pixel; and

repeating said extracting, said obtaining, and said modifying for each of a plurality of destination pixels in said corrected image.

13. The computer-readable storage medium of claim 9, wherein said storing comprises:

writing said distortion information as a correction image.

14. The computer-readable storage medium of claim 13, wherein said correction image comprises a plurality of correctional pixel data.

15. The computer-readable storage medium of claim 14, wherein said correctional pixel data comprises position translation information.

16. The computer-readable storage medium of claim 15, wherein said processing comprises:

obtaining a source pixel value from said captured image;

translating, with reference to said position translation information, said source pixel value from a first location in said captured image to a second location in a corrected image; and

repeating said obtaining and said translating for a plurality of destination pixels in said corrected image.

17. A system comprising:

a lens system;

a buffer operable to store a calibration texture; and

a texturing engine operable to generate pixel data, wherein said texturing engine comprises a plurality of pixel shaders and is configured to access said calibration texture from said buffer, wherein said texturing engine is further operable to use said calibration texture to generate a corrected image from a captured image, wherein said calibration texture is associated with distortion at a plurality of focal lengths, and wherein said calibration texture is further associated with color distortion, wherein said calibration texture is operable to correct a distortion introduced by said lens system, and wherein each of said plurality of shaders is configured to calculate a corrected pixel value for each pixel in said corrected image based on a correction value for each pixel obtained from said calibration texture, and wherein said texturing engine is operable to program said plurality of pixel shaders to use said calibration texture to output each pixel in a proper location in said corrected image.

18. The system of claim 17, wherein said plurality of pixel shaders is configured to generate said corrected pixel by obtaining a correction value from said correction image for said corrected pixel, obtaining a source pixel from said captured image, and modifying said source pixel with reference to said correction value to generate said corrected pixel.

19. The system of claim 17 further comprising:

a camera with said lens system operable to provide said captured image, and

wherein said correction image comprises a two-dimensional image.

20. The system of claim 17, wherein said correction image comprises a plurality of separable one-dimensional images.

21. The system of claim 17, wherein said correction image comprises a one-dimensional radial image.

23. The system of claim 17, wherein said plurality of pixel shaders is configured to process a respective plurality of corrected pixels simultaneously to generate said corrected image.

25. A method of reducing image distortion, said method comprising:

determining distortion information associated with a lens system, comprising generating a first equation describing distortion information associated with said lens system at a first focal length, and a second equation describing information associated with said lens system at a second focal length;

storing said distortion information, wherein said distortion information comprises an equation governing a distortion associated with said lens system;

processing, on a pixel by pixel basis, a captured image taken with said lens system, wherein said processing is performed with reference to said distortion information associated with said lens system at each of said first and second focal lengths; and

generating a corrected image comprising reduced image distortion compared to said captured image using a calibration texture, wherein said calibration texture is operable to correct a distortion introduced by said lens system.