IMAGE PROCESSING DEVICE AND IMAGE PROCESSING METHOD

• US 20110176606A1
• Filed: 06/23/2009
• Published: 07/21/2011
• Est. Priority Date: 06/27/2008
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Fuchie, Takaaki
• Granted Patent: US 8,731,052 B2
• Time in Patent Office: Days
• US Class Current: 375/240.05
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock

Abstract

The present invention can suppress the generated code quantity for each image increment so as not to exceeding the target code quantity without having a usage quantizing factor deviate greatly in a sure manner.

An image encoding device (200) determines a basic quantization parameter (QP_MB) serving as a basic quantizing factor to be predicted in the case that the main encoding generated code comes near the target code quantity at the time of encoding an input image (91). The image encoding device (200) encodes the input image (91) for each feedback control increment by executing quantization using an adapted quantization parameter (QPt) based on an average quantization parameter BaseQP serving as a usage quantizing factor determined based on at least the basic quantization parameter (QP_MB). The image encoding device (200) confirms the generated code quantity of the encoded input image (91) for each feedback control increment, and in the case of predicting that the generated code quantity for each image increment will exceed the target code quantity, increases the value of the adapted quantization parameter (QPt) by increasing the average quantization parameter BaseQP.

Claims

1. An image processing device comprising:

a basic quantizing factor determining unit configured to determine a basic quantizing factor regarding which is predicted that the generated code quantity for each image increment when encoding an input image will approximate to the target code quantity for said each image increment;

an encoding unit configured to encode said input image for each feedback control increment by executing quantization using a usage quantizing factor determined based on at least said basic quantizing factor; and

a feedback control unit configured to confirm the generated code quantity of said input image encoded by said encoding unit for said each feedback control increment, and in the case of predicting that the generated code quantity for said each image increment will exceed the target code quantity for said each image increment, increase said usage quantizing factor.

2. The image processing device according to claim 1, wherein said feedback control unit predicts that the generated code quantity for said each image increment will exceed the target code quantity for said each image increment, in the case that the generated code quantity for said each feedback control increment tends to increase, and also there is insufficient leeway for a remaining code quantity obtained by subtracting the generated code quantity of a portion encoded from the target code quantity of a portion encoded with said each image increment.

3. The image processing device according to claim 2, wherein said feedback control unit determines that the generated code quantity for said each feedback control increment tends to increase when the generated code quantity of the last feedback control increment is greater than the target code quantity of the last feedback control increment.

4. The image processing device according to claim 3, wherein said feedback control unit determines that there is insufficient leeway for said remaining code quantity when said remaining code quantity is smaller than an excessive maximum code quantity regarding which there is a possibility that the generated code quantity with the next feedback control increment may exceed the target code quantity.

5. The image processing device according to claim 2, wherein said feedback control unit predicts that the generated code quantity for each image increment will exceed the target code quantity for said each image increment, in the case that the generated code quantity of a portion encoded with said image increment exceeds said target code quantity of this encoded portion, and also the generated code quantity for each image increment exceeds the target code quantity for said each image increment when assuming that a ratio between the target code quantity and the generated code quantity does not change.

6. The image processing device according to claim 5, wherein said feedback control unit determines that the generated code quantity for each image increment exceeds the target code quantity for said each image increment when assuming that a ratio between said target code quantity and the generated coded quantity does not change, in the case that the predictive value of the generated code quantity for each image increment obtained from the ratio between the target code quantity and the generated code quantity of the last feedback control increment exceeds the target code quantity.

7. The image processing device according to claim 6, wherein said feedback control unit predicts whether or not the generated code quantity for said each image increment will exceed the target code quantity for said each image increment, with a latter-half portion of said image unit.

8. The image processing device according to claim 7, wherein said feedback control unit predicts whether or not the generated code quantity for said each image increment will exceed the target code quantity for said each image increment regardless of the position of said image increment, in the case that there is discrepancy between the generated code quantity of a portion encoded with said image unit and the target code quantity of this encoded portion.

9. The image processing device according to claim 1, wherein said feedback control unit determines that there is discrepancy between the generated code quantity of a portion encoded with said image increment and the target code quantity of this encoded portion, in the case that the generated code quantity of a portion encodes with said image increment is smaller than a value obtained by multiplying a tolerable deviating quantity as to the target code quantity of this encoded portion.

10. The image processing device according to claim 1, wherein said feedback control unit increases said usage quantizing factor by one, in the case of predicting that the generated code for said each image increment will exceed the target code quantity for said each image increment.

11. The image processing device according to claim 9, wherein said feedback control unit determines said basic quantizing factor so as to reduce the generated code quantity when encoding said input image smaller than said target code quantity.

12. The image processing device according to claim 11, wherein said tolerable deviating value is made up of a value less than 1.

13. The image processing device according to claim 1, wherein said encoding unit takes said basic quantizing factor determined for said each image increment as an average quantizing factor, and uses a value to which the offset according to activity as to this average quantizing factor is added as said usage quantizing factor;

and wherein said feedback control unit increases said usage quantizing factor by increasing said average quantizing factor.

14. The image processing device according to claim 7, wherein said basic quantizing factor determining unit determines a predictive quantizing factor regarding which is predicted to approximate to a basic quantizing factor based on the generated code quantity calculated by encoding said input image using a quantizing factor covering a wide range, and determines said basic quantizing factor based on the generated code quantity calculated by encoding said input image using this predictive quantizing factor, and a quantizing factor in the vicinity of this predictive quantizing factor.

15. An image processing method comprising:

a basic quantizing factor determining step arranged to determine a basic quantizing factor regarding which is predicted that the generated code quantity for each image increment when encoding an input image will approximate to the target code quantity for said each image increment;

an encoding step arranged to encode said input image for each feedback control increment to generate an encoded stream by executing quantization using a usage quantizing factor determined based on at least said basic quantizing factor; and

a feedback control step arranged to confirm the generated code quantity of said input image encoded in said encoding step for said each feedback control increment, and in the case of predicting that the generated code quantity for said each image increment will exceed the target code quantity for said each image increment, increase said usage quantizing factor.