Enhancing photograph visual quality using texture and contrast data from near infra-red images

• US 8,503,778 B2
• Filed: 05/13/2010
• Issued: 08/06/2013
• Est. Priority Date: 05/14/2009
• Status: Expired due to Fees

• Current Assignee: National University of Singapore
• Original Assignee: National University of Singapore
• Inventors: Sim, Terence Mong Cheng, Zhang, Xiaopeng
• Primary Examiner(s): Patel, Nirav G
• Application Number:
  US12/779,219
• Publication Number:
  US 20100290703A1
• Time in Patent Office: 1,181 Days
• Field of Search: 382/179, 382/172
• US Class Current: 382/172
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/10048 Infrared image
• CPC Class Codes: G06T 2207/20064 Wavelet transform [DWT]
• CPC Class Codes: G06T 5/10 using non-spatial domain filtering
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 5/94 based on local image properties, e.g. for local contrast enhancement

Abstract

Near infra-red images of natural scenes usually have better contrast and contain rich texture details that may not be perceived in visible light photographs. The contrast and rich texture details form a NIR image corresponding to a visible light image are useful for enhancing the visual quality of the visible light image. To enhance the visual quality of a visible light image using its corresponding near infra-red image, a computer-implemented method computes a weight region mask from the visible light image, transfers contrast data and texture data from the near infra-red image to the visible light image guided by the weighted region mask. The contrast data is computed from the low frequency subbands of the visible light image and corresponding infra-red image after a wavelet transform by matching the histogram of gradient magnitude. The texture data is computed from the high frequency subbands of both images after wavelet transform.

Claims

1. A computer-implemented method for enhancing the visual quality of a visible light image using its corresponding near infra-red image, the method comprising:

computing a weighted region mask from the visible light image, the weighted region mask indicating areas of the visible light image to be enhanced and the extent of enhancement;

wavelet transforming the visible light image and the near infra-red image into a plurality of low frequency and high frequency subbands;

computing a contrast transferred subband from the low frequency subband of the visible light image and the low frequency subband of the near infra-red image guided by the weighted region mask;

computing texture transferred subbands from the high frequency subbands of the visible light image and the high frequency subbands of the near infra-red image guided by the weighted region mask; and

generating a quality enhanced visible light image from the contrast transferred subband and the texture transferred subbands.

2. The method of claim 1, further comprising capturing the visible light image and its corresponding near infra-red image simultaneously, wherein each pixel of the visible light image is aligned with the collocated pixel of the near infra-red image.

3. The method of claim 1, wherein calculating the weighted region mask comprises computing the weighted region mask based on saturation and brightness values of the visible light image.

4. The method of claim 1, wherein wavelet transforming the visible light image comprises:

converting the visible light image from a RGB color space to a HSV color space;

generating a channel-V image from the visible light image in HSV color space, the channel-V image using the intensity values of the visible light image in HSV color space; and

decomposing the channel-V image into a low frequency subband and high subband and the texture transferred subbands.

5. The method of claim 4, wherein wavelet transforming the visible light image further comprises calculating histograms of gradient magnitude and histogram of wavelet coefficients of the high frequency subbands of the channel-V image.

6. The method of claim 1, wherein wavelet transforming the near infra-red image comprises decomposing the near infra-red image into a low frequency subband and high frequency subbands in horizontal, vertical and diagonal directions.

7. The method of claim 1, wherein wavelet transforming the near infra-red image further comprises calculating histograms of gradient magnitude and histogram of wavelet coefficients of the high frequency subbands of the near infra-red image.

8. The method of claim 1, wherein computing the contrast transferred subband comprises:

applying bilateral filtering to the average subband of the visible light image to generate a large-scale layer of the visible light image representing the smoothed brightness of the visible light image; and

applying bilateral filtering to the average subband of the near infra-red image to generate a large-scale layer of the near infra-red image representing the smoothed brightness of the near infra-red image.

9. The method of claim 1, wherein computing the contrast transferred subband further comprises:

generating a detail layer of the visible image representing the smoothed texture of the visible light image; and

generating a detail layer of the near infra-red image representing the smoothed texture of the near infra-red image.

10. The method of claim 1, wherein computing the contrast transferred subband further comprises matching histogram of gradient magnitude of the visible light image with histogram of gradient magnitude of the corresponding near infra-red image.

11. The method of claim 10, further comprises generating a contrast transferred subband using the matching result guided by the weighted region mask.

12. The method of claim 1, wherein computing texture transferred subbands comprises combining the high-frequency subbands of the visible image and the corresponding near infra-red image based on the weighted region map.

31. The method of claim 11, wherein combining the high-frequency subbands of the visible image and the corresponding near infra-red image comprises alpha blending the high-frequency subbands guided by the weighted region mask.

13. A computer program product for enhancing the visual quality of a visible light image using its corresponding near infra-red image, the computer program product comprising a non-transitory computer-readable storage medium containing computer program code for performing the operations:

computing a weighted region mask from the visible light image, the weighted region mask indicating areas of the visible light image to be enhanced and the extent of enhancement;

wavelet transforming the visible light image and the near infra-red image into a plurality of low frequency and high frequency subbands;

computing a contrast transferred subband from the low frequency subband of the visible light image and the low frequency subband of the near infra-red image guided by the weighted region mask;

computing texture transferred subbands from the high frequency subbands of the visible light image and the high frequency subbands of the near infra-red image guided by the weighted region mask; and

generating a quality enhanced visible light image from the contrast transferred subband and the texture transferred subbands.

14. The computer program product of claim 13, further comprising computer program code for capturing the visible light image and its corresponding near infra-red image simultaneously, wherein each pixel of the visible light image is aligned with the collocated pixel of the near infra-red image.

15. The computer program product of claim 13, wherein the computer program code for calculating the weighted region mask comprises computer program code for computing the weighted region mask based on saturation and brightness values of the visible light image.

16. The computer program product of claim 13, wherein the computer program code for wavelet transforming the visible light image comprises computer program code for:

converting the visible light image from a RGB color space to a HSV color space;

generating a channel-V image from the visible light image in HSV color space, the channel-V image using the intensity values of the visible light image in HSV color space; and

decomposing the channel-V image into a low frequency subband and high subband and the texture transferred subbands.

17. The computer program product of claim 16, wherein the computer program code for wavelet transforming the visible light image further comprises computer program code for calculating histograms of gradient magnitude and histogram of wavelet coefficients of the high frequency subbands of the channel-V image.

18. The computer program product of claim 13, wherein the computer program code for wavelet transforming the near infra-red image comprises computer program code for decomposing the near infra-red image into a low frequency subband and high frequency subbands in horizontal, vertical and diagonal directions.

19. The computer program product of claim 13, wherein the computer program code for wavelet transforming the near infra-red image further comprises computer program code for calculating histograms of gradient magnitude and histogram of wavelet coefficients of the high frequency subbands of the near infra-red image.

20. The computer program product of claim 13, wherein the computer program code for computing the contrast transferred subband comprises computer program code for:

applying bilateral filtering to the average subband of the visible light image to generate a large-scale layer of the visible light image representing the smoothed brightness of the visible light image; and

applying bilateral filtering to the average subband of the near infra-red image to generate a large-scale layer of the near infra-red image representing the smoothed brightness of the near infra-red image.

21. The computer program product of claim 13, wherein the computer program code for computing the contrast transferred subband further comprises computer program code for:

generating a detail layer of the visible light image representing the smoothed texture of the visible image; and

generating a detail layer of the near infra-red image representing the smoothed texture of the near infra-red image.

22. The computer program product of claim 13, wherein the computer program code for computing the contrast transferred subband further comprises computer program code for matching histogram of gradient magnitude of the visible light image with histogram of gradient magnitude of the corresponding near infra-red image guided by the weighted region mask.

23. The computer program product of claim 22, further comprises computer program code for generating a contrast transferred subband using the matching result.

24. The computer program product of claim 13, wherein computer program code for computing texture transferred subbands comprises computer program code for combining the high-frequency subbands of the visible light image and the corresponding near infra-red image guided by the weighted region mask.

25. A computer system for enhancing a visual quality of a visible light image using its corresponding near infra-red image, the computer system comprising:

a dual-camera sub-system configured to capture the visible light image and its corresponding near infra-red image simultaneously, the dual-camera sub-system comprising;

a first camera configured to capture the visible light image of a scene;

a second camera configured to capture the corresponding near infra-red image of the scene;

a hot mirror configured to filter natural light received by the dual-camera sub-system; and

an optical remote controller configured to trigger the first camera and the second camera simultaneously; and

a visual quality enhancement sub-system configured to enhance the visual quality of the visible light image using its corresponding near infra-red image, the visual quality enhancement sub-system coupled to the dual-camera sub-system.

26. The computer system of claim 25, wherein the hot mirror is further configured to reflect near infra-red light of the natural light.

27. The computer system of claim 25, wherein the hot mirror is a 45°

dielectric mirror, configured to reflect near infra-red light with an angle of 45°

.

28. The computer system of claim 27, wherein the dielectric mirror is further configured not to block any visible light of the natural light.

29. The computer system of claim 25, wherein the optical remote controller is further configured to set up the first camera and second camera optically aligned.

30. The computer system of claim 25, wherein the optical remote controller is further configured to use a camera setting shared by the first camera and the second camera.