Block noise detecting method and apparatus, and block noise reducing method and apparatus

• US 20070223835A1
• Filed: 08/01/2006
• Published: 09/27/2007
• Est. Priority Date: 03/27/2006
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Nakagata, Shohei, Yamada, Kohji
• Granted Patent: US 7,668,392 B2
• Time in Patent Office: Days
• US Class Current: 382/268
• CPC Class Codes: H04N 19/117 Filters, e.g. for pre-processing or post-processing sub-band filter banks H04N19/635
• CPC Class Codes: H04N 19/14 Coding unit complexity, e.g. amount of activity or edge presence estimation H04N19/146 takes precedence
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/60 using transform coding
• CPC Class Codes: H04N 19/86 involving reduction of coding artifacts, e.g. of blockiness

Abstract

A pixel boundary difference detector detects a pixel boundary difference across a pixel boundary on the basis of a difference value between pixel values of adjacent pixels across the pixel boundary on a picture plane of an input image frame, and a predicted pixel value on the pixel boundary predicted from at least a plurality of pixels on one side of the pixel boundary, an accumulator accumulates the pixel boundary differences for one image frame, and a block noise detector detects a block noise on the basis of a result of the accumulation. The accuracy of block noise detection is improved, the effect of block noise reduction is improved, and the picture quality is improved.

Claims

1. A block noise detecting method comprising the steps of:

detecting a pixel boundary difference across a pixel boundary on the basis of a difference value between pixel values of adjacent pixels across the pixel boundary on a picture plane of an input image frame, and a predicted pixel value on the pixel boundary predicted from at least a plurality of pixels on one side of the pixel boundary; and

accumulating the pixel boundary differences for one image frame to detect a block noise.

2. The block noise detecting method according to claim 1, wherein the difference value between pixel values of the adjacent pixels is compared with a difference value between the predicted pixel value on the pixel boundary predicted from the plural pixels on one side of the pixel boundary and the predicted pixel value on the pixel boundary predicted from a plurality of pixels on the other side of the pixel boundary, and a smaller difference value is detected as a pixel boundary difference across the pixel border.

3. The block noise detecting method according to claim 1, wherein the pixel boundary differences are accumulated in a cycle of a block noise size.

4. The block noise detecting method according to claim 1, wherein the predicted pixel value on the pixel boundary is calculated from at least two pixels on one side of the pixel border.

5. The block noise detecting method according to claim 1, wherein when the pixel boundary difference is larger than a predetermined threshold value, the pixel border difference is excluded from a group of pixel boundary differences to be accumulated.

6. The block noise detecting method according to claim 1, wherein ratios each of which is a ratio of an accumulated value obtained by accumulating pixel boundary differences across a pixel boundary for one image frame to an accumulated value obtained by accumulating pixel boundary differences across a pixel boundary separated by a half of a block noise size from the former pixel boundary are compared with one another to determine a block noise generation intensity.

7. The block noise detecting method according to claim 1, wherein a block noise generation intensity is determined on the basis of a ratio of the largest value of accumulated values each of which is obtained by accumulating pixel boundary differences across a pixel boundary for one image frame to the second largest value of the accumulated values.

8. The block noise detecting method according to claim 1, wherein a difference between at least two image frames including the input image frame is determined, and determination conditions of block noise detection is controlled such that the block noise is difficult to be detected when a value of the difference is smaller than a predetermined threshold value.

9. The block noise detecting method according to claim 1, wherein when the block noise is detected on an image enlarged by a scaling factor of an integer multiple after the image is decoded, pixel boundary differences are accumulated in a cycle corresponding to the integer multiple of a block noise size.

10. The block noise detecting method according to claim 1, wherein when the same value is obtained as a result of detection of the block noise from successive plural image frames, this result of the detection is outputted.

11. A block noise detecting apparatus comprising:

a pixel boundary difference detecting means for detecting a pixel boundary difference across a pixel boundary on the basis of a difference value between pixel values of adjacent pixels across the pixel boundary on a picture plane of an input image frame, and a predicted pixel value on the pixel boundary predicted from at least a plurality of pixels on one side of the pixel boundary;

an accumulating means for accumulating the pixel boundary differences for one image frame; and

a block noise detecting means for detecting a block noise on the basis of a result of accumulation by said accumulating means.

12. The block noise detecting apparatus according to claim 11, wherein said pixel boundary difference detecting means comprises:

an adjacent pixel difference value calculating unit for determining the difference value between pixel values of the adjacent pixels;

a predicted pixel difference value calculating unit for determining a difference value between the predicted pixel value on the pixel boundary predicted from a plurality of pixels on one side of the pixel boundary and the predicted value on the pixel boundary predicted from a plurality of pixels on the other side of the pixel boundary; and

a comparing unit for comparing results of calculation by the calculating units, and detecting a smaller difference value as a pixel boundary difference across the pixel boundary.

13. The block noise detecting apparatus according to claim 11, wherein said accumulating means accumulates the pixel boundary differences in a cycle of a block noise size.

14. The block noise detecting apparatus according to claim 11, wherein said predicted pixel difference value calculating unit calculates the predicted pixel value on the pixel boundary from at least two pixels on one side of the pixel boundary.

15. The block noise detecting apparatus according to claim 11, wherein when the pixel boundary difference is larger than a predetermined threshold value, said pixel boundary difference detecting means or said accumulating means excludes the pixel boundary difference from a group of pixel boundary differences to be accumulated.

16. The block noise detecting apparatus according to claim 11, wherein said block noise detecting means comprises:

a means for comparing, with one another, ratios each of which is a ratio of an accumulated value obtained by accumulating pixel border differences across a pixel border for one image frame by said accumulating means to an accumulated value obtained by accumulating pixel border differences across a pixel border separated by a half of a block noise size from the former pixel boundary to determine a block noise generation intensity.

17. The block noise detecting apparatus according to claim 11, wherein said block noise detecting means comprises:

a means for determining a block noise generation intensity on the basis of a ratio of the largest value of accumulated values each of which is obtained by accumulating pixel border differences across a pixel boundary for one image frame by said accumulating means with the second largest value of the accumulated values.

18. The block noise detecting apparatus according to claim 11, further comprising:

a frame difference detecting means for determining a difference between at least two image frames including the input image frame; and

a controlling means for, when a value of the difference detected by said frame difference detecting means is smaller than a predetermined threshold value, controlling determination conditions of block noise detection by said block noise detecting means such that the block noise is difficult to be detected.

19. The block noise detecting apparatus according to claim 11, wherein when the block noise is detected on an image enlarged by a scaling factor of an integer multiple after the image is decoded, said accumulating means accumulates the pixel border differences in a cycle corresponding to the integer multiple of a block noise size.

20. The block noise detecting apparatus according to claim 11, further comprising:

a block noise detection result inspecting means for, when the same value is obtained as a result of detection by said block noise detecting means from successive plural image frames, outputting this result of the detection.

21. A block noise reducing method comprising the steps of:

detecting a pixel boundary difference across a pixel boundary on the basis of a difference value between pixel values of adjacent pixels across the pixel boundary on a picture plane of an input image frame, and a predicted pixel value on the pixel boundary predicted from at least a plurality of pixels on one side of the pixel boundary;

accumulating the pixel boundary differences for one image frame to detect a block noise; and

performing a block noise reducing process on the input image frame on the basis of the detected block noise.

22. The block noise reducing method according to claim 21, wherein a filtering intensity of a filtering process as being the block noise reducing process is controlled on the basis of ratios each of which is a ratio of an accumulated value obtained by accumulating pixel border differences across a pixel boundary for one image frame to an accumulated value obtained by accumulating pixel border differences across a pixel boundary separated by a half of a block noise size from the former pixel boundary.

23. The block noise reducing method according to claim 21, wherein a filtering intensity of a filtering process as being the block noise reducing process is controlled on the basis of a ratio of the largest value of accumulated values each of which is obtained by accumulating pixel boundary differences across a pixel boundary for one image frame to the second largest value of the accumulated values.

24. The block noise reducing method according to claim 21, wherein a region on the picture plane to which the block noise reducing process is applied is limited to a pixel boundary and vicinity thereof where a block noise is generated on the basis of the result of block noise detection.

25. The block noise reducing method according to claim 21, wherein when the pixel boundary difference is equal to or larger than a threshold value, the block noise reducing process is not performed.

26. A block noise reducing apparatus comprising:

a pixel boundary difference detecting means for detecting a pixel boundary difference across a pixel boundary on the basis of a difference value between pixel values of adjacent pixels across the pixel boundary on a picture plane of an input image frame, and a predicted pixel value on the pixel boundary predicted from at least a plurality of pixels on one side of the pixel boundary;

an accumulating means for accumulating the pixel boundary differences for one image frame;

a block noise detecting means for detecting a block noise on the basis of a result of accumulation by said accumulating means; and

a block noise reducing means for performing a block noise reducing process on the input picture frame on the basis of the block noise detected by said block noise detecting means.

27. The block noise reducing apparatus according to claim 26, further comprising:

a filter controlling means for controlling a filtering intensity of a filtering process as being the block noise reducing process by said block noise reducing means on the basis of ratios each of which is a ratio of an accumulated value obtained by accumulating pixel boundary differences across a pixel boundary for one image frame by said accumulating means to an accumulated value obtained by accumulating pixel boundary differences across a pixel boundary separated a half of a block noise size from the former pixel boundary.

28. The block noise reducing apparatus according to claim 26, further comprising:

a filter controlling means for controlling a filtering intensity of a filtering process as being the block noise reducing process by said block noise reducing means on the basis of a ratio of the largest value of accumulated values each of which is obtained by accumulating pixel boundary differences across a pixel boundary for one image frame by said accumulating means to the second largest value of the accumulated values.

29. The block noise reducing apparatus according to claim 26, further comprising:

a filtering region controlling means for limiting a region on the picture plane to which the filtering process is applied by said block noise reducing means to a pixel boundary and vicinity thereof where a block noise is generated on the basis of a result of detection by said block noise detecting means.

30. The block noise reducing apparatus according to claim 26, further comprising:

a filter interruption controlling means for controlling to interrupt the filtering process by said block noise reducing means when the pixel boundary difference is equal to or larger than a threshold value.