Image processing apparatus and method

• US 8,184,182 B2
• Filed: 10/06/2009
• Issued: 05/22/2012
• Est. Priority Date: 11/19/2008
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Lee, Seok, Wey, Ho-cheon
• Primary Examiner(s): Tran, Nhan T
• Application Number:
  US12/574,181
• Publication Number:
  US 20100123807A1
• Time in Patent Office: 959 Days
• Field of Search: 348/241, 348/222.1, 348/208.99, 348/208.4, 348/208.7, 348/208.13, 348/273, 348/234, 348236-238, 348/362, 348/229.1, 348/297, 382/254, 382/275, 382/279
• US Class Current: 348/241
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 2207/20201 Motion blur correction
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening

Abstract

An image processing apparatus and method are provided. The image processing apparatus includes a blur kernel estimator and an image restorer. The blur kernel estimator is configured to estimate a blur kernel of a first image using the first image and a second image, wherein the first image includes multi-channel color image data and the second image includes single-channel image data and is obtained with a shorter exposure time than the first image. The image restorer is configured to generate a blurred image of the second image using the blur kernel, and restore an image using residual images between images for respective channels of the first image and the blurred image of the second image, the second image, and the blur kernel.

Claims

1. An image processing apparatus, comprising:

a blur kernel estimator configured to estimate a blur kernel of a first image using the first image and a second image, wherein

the first image comprises multi-channel color image data, and

the second image comprises single-channel image data and is obtained with a shorter exposure time than the first image; and

an image restorer configured to generate a blurred image of the second image using the blur kernel, and restore an image using residual images between images for respective channels of the first image and the blurred image of the second image, the second image, and the blur kernel.

2. The image processing apparatus of claim 1, wherein:

the first image is an image detected from a light signal belonging to a color band of an optical region among incident light signals, and

the second image is an image detected from at least one or combination of a white signal, a white signal including infrared rays, an infrared signal, and a light signal belonging to a complementary wavelength band among the incident light signals.

3. The image processing apparatus of claim 1, wherein:

the blur kernel estimator comprises a channel unification processor configured to perform a channel unification, wherein the channel unification processor converts, in response to receiving a multi-channel image signal, the multi-channel image signal into a gray-scale image, generates edge images of images for respective channels corresponding to the multi-channel image signal, and synthesizes the gray-scale image with the edge images to generate a single-channel image, and

the first image is subjected to the channel unification by the channel unification processor to estimate the blur kernel.

4. The image processing apparatus of claim 3, wherein the channel unification processor generates the second image by performing the channel unification on a multi-channel image obtained with a shorter exposure time than the first image.

5. The image processing apparatus of claim 3, wherein the residual images have a smaller signal magnitude than the first image subjected to the channel unification.

6. The image processing apparatus of claim 1, wherein the residual images are generated by subtracting the blurred image of the second image from the images for the respective channels of the first image.

7. The image processing apparatus of claim 6, wherein the image restorer generates deblurred images for the residual images for the respective channels by performing deconvolution on the residual images with the blur kernel, and restores the image by adding the second image to the deblurred images.

8. The image processing apparatus of claim 7, wherein the image restorer performs deconvolution based on edge data extracted from the second image.

9. The image processing apparatus of claim 8, wherein the image restorer performs deconvolution on an image with weights assigned to respective regions of the image.

10. The image processing apparatus of claim 1, further comprising an image obtainer configured to obtain the first and second images.

11. The image processing apparatus of claim 1, wherein the second image is an image obtained with higher sensitivity than the first image.

12. The image processing apparatus of claim 1, wherein the second image is of the same scene as the first image and is an image obtained from optical signals of a wide-band wavelength.

13. The image processing apparatus of claim 1, wherein the blur kernel estimator comprises:

a channel unification processor configured to keep edge data of the first image while the first image is subjected to channel unification; and

a blur kernel extractor configured to extract the blur kernel from the channel-unified first image and the second image.

14. An image processing method performed by an image processing apparatus, the method comprising:

obtaining a first image having multi-channel color image data, and a second image having single-channel image data obtained with a shorter exposure time than the first image;

estimating a blur kernel of the first image using the first image and the second image;

generating a blurred image of the second image using the blur kernel;

generating residual images for respective channels of the first image between images for the respective channels and the blurred image of the second image; and

restoring an image using the second image, the residual images, and the blur kernel.

15. The image processing method of claim 14, wherein:

the first image is an image detected from a light signal belonging to a color band of the optical region among incident light signals, and

the second image is an image detected from at least one or combination of a white signal, a white signal including infrared rays, an infrared signal, and a light signal belonging to a complementary wavelength band among the incident light signals.

16. The image processing method of claim 14, wherein:

the estimating of the blur kernel comprises performing channel unification on the first image, and

the performing of the channel unification comprises;

converting the first image into a gray-scale image, and generating edge images of the images for the respective channels of the first image; and

generating a single-channel image by synthesizing the gray-scale image and the edge images.

17. The image processing method of claim 16, further comprising generating the second image by performing the channel unification on multi-channel image data obtained with a shorter exposure time than the first image.

18. The image processing method of claim 14, wherein the generating of the residual images comprises generating the residual images by subtracting the blurred image of the second image from the images for the respective channels of the first image.

19. The image processing method of claim 18, wherein the restoring of the image comprises:

generating deblurred images of the residual images by performing deconvolution on the residual images with the blur kernel; and

restoring the image by adding the second image to the deblurred images.

20. The image processing method of claim 19, wherein the restoring of the image comprises performing deconvolution based on edge data extracted from the second image.

21. A non-transitory computer-readable storage medium storing a program to process an image, comprising instructions to cause a computer to:

obtain a first image having multi-channel color image data, and a second image having single-channel image data obtained with a shorter exposure time than the first image;

estimate a blur kernel of the first image using the first image and the second image;

generate a blurred image of the second image using the blur kernel;

generate residual images for respective channels of the first image between images for the respective channels and the blurred image of the second image; and

restore an image using the second image, the residual images, and the blur kernel.

22. The computer-readable storage medium of claim 21, wherein the instruction to generate the residual images comprises an instruction to generate the residual images by subtracting the blurred image of the second image from the images for the respective channels of the first image.