METHOD AND APPARATUS FOR IMAGE PROCESSING

• US 20150213607A1
• Filed: 01/23/2015
• Published: 07/30/2015
• Est. Priority Date: 01/24/2014
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: JEONG, Young Ju, KIM, Ji Won, NAM, Dong Kyung, PARK, Du-Sik, CHO, Yang Ho, HWANG, Hyo Seok
• Granted Patent: US 9,679,395 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06T 2207/20228 Disparity calculation for image-based rendering
• CPC Class Codes: G06T 7/97 Determining parameters from multiple pictures depth or shape recovery from multiple images G06T7/55; stereo camera calibration G06T7/85

Abstract

At least one example embodiment discloses an imaging apparatus that may receive input images with different viewpoints, obtain information associated with the input images, and generate a new viewpoint image based on the received input images and the obtained information and thus, reduce an information loss of the new viewpoint image for the input images.

Claims

1. An image processing method, comprising:

calculating a motion vector of points corresponding to input images with different viewpoints based on homography information and disparity information on the input images, the homography information being based on calibration information and rectification information on the input images; and

generating a new viewpoint image from the input images based on the motion vector.

2. The method of claim 1, wherein the calibration information comprises rotation information for rectification of the input images.

3. The method of claim 1, wherein the rectification information comprises parameters associated with rectified input images.

4. The method of claim 1, wherein the input images comprise a left input image and a right input image, and

a viewpoint part of the new viewpoint image is between a viewpoint part of the left input image and a viewpoint part of the right input image.

5. The method of claim 1, wherein the input images include a first input image and a second input image,

the homography information includes first homography information on the first input image and second homography information on the second input image, and

the calculating of the motion vector includes,

calculating a second point of a first rectified image corresponding to a first point of the first input image based on the first homography information;

calculating a third point of a second rectified image corresponding to the second point based on the disparity information;

calculating a fourth point of the second input image corresponding to the third point based on the second homography information; and

calculating the motion vector, the motion vector indicating a motion between the first point and the fourth point.

6. The method of claim 1, further comprising:

obtaining the calibration information, the rectification information, and the disparity information.

7. The method of claim 6, wherein the obtaining the calibration information, the rectification information, and the disparity information comprises:

receiving at least one set of information among the calibration information, the rectification information, and the disparity information; and

generating remaining information, excluding the at least one received set of information, based on the at least one received set of information and the input images.

8. The method of claim 6, wherein the obtaining the calibration information, the rectification information, and the disparity information comprises:

receiving the calibration information; and

generating the rectification information by performing rectification on the input images based on the calibration information and the input images.

9. The method of claim 8, wherein the obtaining the calibration information, the rectification information, and the disparity information comprises:

obtaining the disparity information by calculating disparities based on the rectification.

10. The method of claim 8, further comprising:

calculating an information loss of cropped images for the input images,

wherein the cropped images are obtained by cropping performed concurrently with the rectification, and

the calculating the motion vector and the generating the new viewpoint image are performed when the information loss is greater than or equal to a value.

11. A non-transitory computer-readable medium comprising a program for instructing a computer to perform the method of claim 1.

12. An imaging apparatus, comprising:

a motion vector calculator configured to calculate a motion vector of points corresponding to input images with different viewpoints based on homography information and disparity information on the input images, the homography information being based on calibration information and rectification information on the input images; and

an image generator to generate a new viewpoint image from the input images based on the motion vector.

13. The imaging apparatus of claim 12, wherein the calibration information includes rotation information for rectification of the input images.

14. The imaging apparatus of claim 12, wherein the input images include a first input image and a second input image, and

the motion vector calculator is configured to calculate a second point of a first rectified image corresponding to a first point of the first input image based on a part of the homography information corresponding to the first input image, calculate a third point of a second rectified image corresponding to the second point based on the disparity information, calculate a fourth point of the second input image corresponding to the third point based on another part of the homography information corresponding to the second input image, and calculate the motion vector indicating a motion between the first point and the fourth point.

15. The imaging apparatus of claim 12, further comprising:

an information acquirer configured to obtain the homography information on the input images based on the calibration information and the rectification information.

16. The imaging apparatus of claim 15, wherein the information acquirer is configured to obtain the calibration information, the rectification information, and the disparity information, and obtain the homography information based on the obtained calibration information and the obtained rectification information.

17. The imaging apparatus of claim 15, wherein the information acquirer is configured to receive at least one set of information among the calibration information, the rectification information, and the disparity information, and generate remaining information, excluding the at least one received set of information, based on the at least one received set of information and the input images.

18. The imaging apparatus of claim 15, wherein the information acquirer is configured to receive the calibration information and generate the rectification information by performing rectification on the input images based on the calibration information and the input images.

19. The imaging apparatus of claim 18, wherein the information acquirer is configured to obtain the disparity information by calculating disparities of images generated by the rectification.

20. The imaging apparatus of claim 18, wherein the information acquirer is configured to calculate an information loss of cropped images for the input images,

the cropped images based on cropping performed concurrent with the rectification, and

if the information loss is greater than or equal to a value, the motion vector calculator is configured to calculate the motion vector and the image generator is configured to generate the new viewpoint image.