Adaptive image improvement

• US 20060034512A1
• Filed: 07/26/2004
• Published: 02/16/2006
• Est. Priority Date: 07/26/2004
• Status: Active Grant

• Current Assignee: Rateze Remote Mgmt. L.L.C. (Intellectual Ventures LLC)
• Original Assignee: Somle Development LLC (Intellectual Ventures LLC)
• Inventors: Sheraizin, Vitaly S., Sheraizin, Semion M.
• Granted Patent: US 7,903,902 B2
• Time in Patent Office: Days
• US Class Current: 382/173
• CPC Class Codes: H04N 1/4092 Edge or detail enhancement
• CPC Class Codes: H04N 1/628 Memory colours, e.g. skin or sky

Abstract

A method includes analyzing an input image at least to determine locations of human skin in the input image and processing the input image at least to improve, on a per pixel basis, the areas of human skin of the input image. Another method included in the present invention includes measuring blurriness levels in an input image; and processing the input image with the blurriness levels at least to sharpen the input image. A third method includes identifying areas of at least bright light in an input image and changing the sharpness of the input image as a function of exposure level of different areas of the input image.

Claims

1. A system comprising:

an image analyzer to analyze an input image at least to determine locations of human skin in said input image; and

a processing unit, using the output of said image analyzer, at least to improve, on a per pixel basis, said areas of human skin of said input image.

2. The system according to claim 1 and wherein said processing unit comprises:

a skin processing unit to reduce saturation levels for areas of human skin under low light exposures; and

a sharpness controller to reduce sharpness levels for human skin.

3. A system comprising:

an image analyzer to measure blurriness levels in an input image; and

a processing unit, using said blurriness levels, at least to sharpen said input image.

4. The system according to claim 3 and wherein said image analyzer comprises a sharpness detector to indicate which pixels have edges of predefined widths.

5. The system according to claim 4 and wherein said processing unit comprises a coefficient generator to utilize the output of said sharpness detector to generate per pixel sharpness coefficients and at least a multiplier to multiply high frequency components of said image by said per pixel sharpness coefficients.

6. The system according to claim 5 and wherein said per pixel sharpness coefficients are a function of said predefined widths, where wider edges receive larger coefficients.

7. The system according to claim 6 and wherein said per pixel sharpness coefficients are different for locations of bright light.

8. A system comprising:

an image analyzer to identify areas of at least bright light in an input image; and

a processing unit to change the sharpness of said input image as a function of exposure level of different areas of said input image.

9. The system according to claim 8 and wherein said image analyzer comprises a brightness analyzer to generate a bright light mask for pixels with intensities at or above a predefined bright light level.

10. The system according to claim 9 and wherein said processing unit comprises a sharpness coefficient generator to generate per pixel sharpness coefficients for said pixels with intensities at or above said predefined bright light level.

11. A system comprising:

an image analyzer to analyze an input image;

a skin processing unit to reduce saturation levels for areas of human skin under low light exposures; and

a visual resolution enhancer to improve the visual resolution of said input image.

12. The system according to claim 11 and wherein said image analyzer comprises units to determine locations of human skin, amount of texture, duration of edges and locations of bright light and low light in said input image.

13. The system according to claim 12 and wherein said system also comprises a controller to generate a low light skin mask, a texture threshold, per pixel sharpness coefficients, per pixel texture coefficients and per pixel skin coefficients from said output of said image analyzer.

14. The system according to claim 13 and wherein said controller is connected to said skin processing unit to, provide said low light skin mask.

15. The system according to claim 13 and wherein said visual resolution enhancer comprises:

a sharpness enhancer to enhance sharpness of high frequency components of distinguishable small details using said per pixel sharpness coefficients; and

a sharpness controller to reduce sharpness levels for areas of human skin using said per pixel skin coefficients.

16. The system according to claim 15 and wherein said sharpness enhancer comprises a non-linear operator to reduce the intensity levels of said high frequency components of distinguishable small details, wherein said non-linear operator reduces high levels of said high frequency components more strongly than it reduces lower levels of said high frequency components.

17. The system according to claim 15 and wherein said visual resolution enhancer also comprises:

a texture limiter to limit high frequency components of said image to be no higher than said texture threshold; and

a texture modifier to modify said limited high frequency components using said per pixel texture coefficients.

18. The system according to claim 15 and wherein said visual resolution enhancer also comprises a noise reducer and wherein said noise reducer comprises a first low pass filter before said sharpness enhancer and a texture low pass filter after said texture modifier.

19. The system according to claim 18 and wherein said visual resolution enhancer comprises a horizontal high frequency channel and a vertical high frequency channel and wherein said texture limiter, texture modifier and texture low pass filter operate on both said horizontal and vertical channels.

20. The system according to claim 19 and wherein said texture low pass filter comprises a horizontal low pass filter in said vertical channel and a vertical low pass filter in said horizontal channel.

21. A method comprising:

analyzing an input image at least to determine locations of human skin in said input image; and

processing said input image, using the results of said analyzing, at least to improve, on a per pixel basis, said areas of human skin of said input image.

22. The method according to claim 21 and wherein said processing comprises:

reducing saturation levels for areas of human skin under low light exposures; and

reducing sharpness levels for human skin.

23. A method comprising:

measuring blurriness levels in an input image; and

processing said input image, using said blurriness levels, at least to sharpen said input image.

24. The method according to claim 23 and wherein said measuring comprises indicating which pixels have edges of predefined widths.

25. The method according to claim 24 and wherein said processing comprises generating per pixel sharpness coefficients utilizing the output of said measuring and at least multiplying high frequency components of said image by said per pixel sharpness coefficients.

26. The method according to claim 25 and wherein said per pixel sharpness coefficients are a function of said predefined widths, where wider edges receive larger coefficients.

27. The method according to claim 26 and wherein said per pixel sharpness coefficients are different for locations of bright light.

28. A method comprising:

identifying areas of at least bright light in an input image; and

changing the sharpness of said input image as a function of exposure level of different areas of said input image.

29. The method according to claim 28 and wherein said identifying comprises generating a bright light mask for pixels with intensities at or above a predefined bright light level.

30. The method according to claim 29 and wherein said changing comprises generating per pixel sharpness coefficients for said pixels with intensities at or above said predefined bright light level.

31. A method comprising:

analyzing an input image;

reducing saturation levels for areas of human skin under low light exposures; and

improving the visual resolution of said input image.

32. The method according to claim 31 and wherein said analyzing comprises determining locations of human skin, amount of texture, duration of edges and locations of bright light and low light in said input image.

33. The method according to claim 32 and also comprising generating a low light skin mask, a texture threshold, per pixel sharpness coefficients, per pixel texture coefficients and per pixel skin coefficients from said output of said analyzing.

34. The method according to claim 33 and wherein said generating comprises providing said low light skin mask to said step of reducing.

35. The method according to claim 33 and wherein said improving comprises:

enhancing sharpness of high frequency components of distinguishable small details using said per pixel sharpness coefficients; and

reducing sharpness levels for areas of human skin using said per pixel skin coefficients.

36. The method according to claim 35 and wherein said enhancing comprises reducing the intensity levels of said high frequency components of distinguishable small details by reducing high levels of said high frequency components more strongly than it reduces lower levels of said high frequency components.

37. The method according to claim 35 and wherein said improving also comprises:

limiting high frequency components of said image to be no higher than said texture threshold; and

modifying said limited high frequency components using said per pixel texture coefficients.

38. The method according to claim 35 and wherein said improving also comprises low pass filtering before said enhancing and texture low pass filtering after said modifying.

39. The method according to claim 38 and wherein said improving comprises a horizontal high frequency channel and a vertical high frequency channel and wherein said limiting, modifying and texture low pass filtering operate on both said horizontal and vertical channels.

40. The method according to claim 39 and wherein said texture low pass filtering comprises horizontal low pass filtering in said vertical channel and a vertical low pass filtering in said horizontal channel.