Image Processing Method, Apparatus, and Computer Program Product

• US 20120128256A1
• Filed: 03/16/2011
• Published: 05/24/2012
• Est. Priority Date: 11/19/2010
• Status: Active Grant

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Mishima, Nao, Mita, Takeshi
• Granted Patent: US 8,682,060 B2
• Time in Patent Office: Days
• US Class Current: 382/209
• CPC Class Codes: G06T 7/254 involving subtraction of images

Abstract

According to one embodiment, an image processing method includes calculating an activity value for a first block, the activity value indicating a higher degree of activity as pixel values vary in a greater degree in the first block, and calculating a first evaluation value that indicates higher evaluation as a difference between a pixel value of the first block and a pixel value of a second block is smaller. The method further includes calculating a second evaluation value that indicates higher evaluation as correlation between a relative spatial relationship of the pixel value of the first block and that of the pixel value of the second block is higher, and calculating a third evaluation value by weighting the first evaluation value and the second evaluation value to search for the second block that corresponds to the first block, a weight of the first evaluation value is larger as the activity value is larger.

Claims

1. An image processing method comprising:

calculating an activity value for a first block that is an image region that includes part of pixels of a first image, the activity value indicating a higher degree of activity as pixel values vary in a greater degree among the pixels in the first block;

calculating a first evaluation value that indicates higher evaluation as a difference between a pixel value of the first block and a pixel value of a second block that is an image region including part of pixels of a second image is smaller;

calculating a second evaluation value that indicates higher evaluation as correlation between a relative spatial relationship of the pixel value of the first block and a relative spatial relationship of the pixel value of the second block is higher; and

calculating a third evaluation value by weighting the first evaluation value and the second evaluation value to search for the second block that corresponds to the first block, a weight of the first evaluation value is larger as the activity value is larger.

2. The image processing method according to claim 1, wherein, in calculating the activity value, the activity value that indicates the higher degree for a greater standard deviation of the pixel values of the first block is calculated.

3. The image processing method according to claim 1, wherein, in calculating the activity value, the activity value that indicates the higher degree for a greater mean absolute difference of the pixel values of adjacent pixels in the first block is calculated.

4. The image processing method according to claim 1, wherein, in calculating the third evaluation value, the third evaluation value is calculated by changing the weight in accordance with an increase function incorporating the activity value as a variable.

5. The image processing method according to claim 1, wherein, in calculating the second evaluation value, the second evaluation value that indicates higher evaluation for a higher correlation between a relative relationship of the pixel values of adjacent pixels in the first block and a relative relationship of the pixel values of adjacent pixels in the second block is calculated.

6. The image processing method according to claim 1, wherein, in calculating the second evaluation value, the second evaluation value that indicates higher evaluation for a higher correlation between a relative relationship of the pixel values of two pixels or more aligned in a horizontal direction in the first block and a relative relationship of the pixel values of two pixels or more aligned in the horizontal direction in the second block is calculated.

7. The image processing method according to claim 1, wherein, in calculating the second evaluation value, the second evaluation value that indicates higher evaluation for a higher correlation between a relative relationship of the pixel values of two pixels or more aligned in a vertical direction in the first block and a relative relationship of the pixel values of two pixels or more aligned in the vertical direction in the second block is calculated.

8. The image processing method according to claim 1, wherein, in calculating the first evaluation value, the first evaluation value that is at least any one of a mean absolute difference, a sum of absolute differences, a mean squared error, and a sum of squared differences is calculated by use of the pixel value of the first block and the pixel value of the second block.

9. The image processing method according to claim 1, wherein, in calculating the second evaluation value, the second evaluation value that is at least either one of an increment sign correlation and a difference between gradients of the pixel values is calculated by use of the pixel value of the first block and the pixel value of the second block.

10. The image processing method according to claim 1, further comprising searching for the second block, in which the third evaluation value is maximized, as a block that corresponds to the first block.

11. An image processing apparatus comprising:

a first calculating unit that calculates an activity value for a first block that is an image region that includes part of pixels of a first image, the activity value indicating a higher degree of activity as pixel values vary in a greater degree among the pixels in the first block;

a second calculating unit that calculates a first evaluation value that indicates higher evaluation as a difference between the pixel values of the first block and pixel values of a second block that is an image region including part of a second image is smaller;

a third calculating unit that calculates a second evaluation value that indicates higher evaluation as correlation between a relative spatial relationship of the pixel values of the first block and a relative spatial relationship of the pixel values of the second block is higher; and

a fourth calculating unit that calculates a third evaluation value by weighting the first evaluation value and the second evaluation value to search for the second block that corresponds to the first block, a weight of the first evaluation value is larger as the activity value is larger.

12. A computer program product having a non-transitory computer readable medium including programmed instructions, wherein the instructions, when executed by a computer, causes the computer to execute:

calculating an activity value for a first block that is an image region that includes part of pixels of a first image, the activity value indicating a higher degree of activity as pixel values vary in a greater degree among the pixels in the first block;

calculating a first evaluation value that indicates higher evaluation as a difference between a pixel value of the first block and a pixel value of a second block that is an image region including part of pixels of a second image is smaller;

calculating a second evaluation value that indicates higher evaluation as correlation between a relative spatial relationship of the pixel value of the first block and a relative spatial relationship of the pixel value of the second block is higher; and

calculating a third evaluation value by weighting the first evaluation value and the second evaluation value to search for the second block that corresponds to the first block, a weight of the first evaluation value is larger as the activity value is larger.