Encoding Method, Encoding Apparatus, Decoding Method, Decoding Apparatus and Program

• US 20050261893A1
• Filed: 06/11/2002
• Published: 11/24/2005
• Est. Priority Date: 06/15/2001
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Tsuji, Minoru, Nishiguchi, Masayuki, Toyama, Keisuke, Suzuki, Shiro
• Granted Patent: US 7,212,973 B2
• Time in Patent Office: Days
• US Class Current: 704/201
• CPC Class Codes: G10L 19/02 using spectral analysis, e.g. transform vocoders or subband vocoders
• CPC Class Codes: G10L 19/032 Quantisation or dequantisation of spectral components
• CPC Class Codes: G10L 19/06 Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients
• CPC Class Codes: H03M 7/30 Compression speech analysis-synthesis for redundancy reduction G10L19/00; for image communication H04N; Expansion; Suppression of unnecessary data, e.g. redundancy reduction

Abstract

In a quantization step information encoding unit, an average value of the quantization step information is found in an approximate shape extraction unit (20), first of all, from one set of a given number of unitary quantization units to another. In an approximate shape encoding unit (21), the approximate shape information is vector-quantized. In a residual signal computing unit (22), the residual signals between the quantization step information and the quantized approximate shape vector are computed. In a residual signal encoding unit (23), the residual signals are variable length encoded, and the so encoded residual signals and the vector quantized approximate shape information are output.

Claims

1. An encoding apparatus for encoding the main information of audio and/or video signals, which is the direct subject-matter of encoding, and the subsidiary information, which is not the direct subject-matter of encoding, said apparatus comprising:

approximate shape computing means for computing the approximate shape information of said subsidiary information;

approximate shape encoding means for quantizing and encoding said approximate shape information;

error computing means for computing the quantization error in said approximate shape encoding means; and

error encoding means for encoding said quantization error in a lossless fashion without losing the information;

wherein

said approximate shape information encoded by said approximate shape encoding means and said quantization error encoded by said error encoding means are output.

2. The encoding apparatus according to claim 1 wherein

said approximate shape computing means finds the approximate shape of said subsidiary information by linear prediction coefficients of said subsidiary information.

3. The encoding apparatus according to claim 1 wherein

said approximate shape computing means finds the approximate shape of said subsidiary information by cepstrum coefficients of said subsidiary information.

4. The encoding apparatus according to claim 1 wherein

said approximate shape computing means computes a representative value of said subsidiary information with respect to each preset unitary unit to find an approximate shape of said subsidiary information.

5. The encoding apparatus according to claim 4 wherein

said representative value is an average value.

6. The encoding apparatus according to claim 4 wherein

said representative value is a maximum value.

9. The encoding apparatus according to claim 1 wherein

said approximate shape computing means computes the difference between said subsidiary information and the approximate shape information encoded by said approximate shape encoding means.

10. The encoding apparatus according to claim 1 wherein

said error encoding means encodes said quantization error into variable length codewords.

11. The encoding apparatus according to claim 10 wherein

said variable length codewords are composed of Huffman codes.

12. The encoding apparatus according to claim 1 wherein

said subsidiary information is the normalization information.

13. The encoding apparatus according to claim 1 wherein

said subsidiary information is the quantization step information.

14. The encoding apparatus according to claim 1 further comprising:

splitting means for splitting said audio and/or video signals into preset encoding units;

decision means for deciding on the normalization information and the quantization step information, from one encoding unit to another;

normalization means for normalizing said audio and/or video signals based on said normalization information;

quantization means for quantizing said audio and/or video signals based on said quantization step information; and

encoding means for encoding said audio and/or video signals in accordance with the encoding system associated with that for said quantization step information.

15. The encoding apparatus according to claim 1 further comprising:

comparator means for comparing the code volume by said error encoding means to the code volume by another encoding system for encoding said subsidiary information; and

selection means for selecting the code volume which has been found to be smaller on comparison by said comparator means.

7. The encoding apparatus according to claim I wherein

said approximate shape encoding means vector-quantizes the approximate shape information as computed by said approximate shape computing means.

8. The encoding apparatus according to claim 7 wherein

said approximate shape encoding means encodes such an index of a codebook, provided at the outset for encoding the approximate shape information, which index will minimize the quantization error.

16. An encoding method for encoding the main information of audio and/or video signals, which is the direct subject-matter of encoding, and the subsidiary information, which is not the direct subject-matter of encoding, said method comprising:

an approximate shape computing step of computing the approximate shape information of said subsidiary information;

an approximate shape encoding step of quantizing and encoding said approximate shape information;

an error computing step of computing the quantization error; and

an error encoding step of encoding said quantization error in a lossless fashion without losing the information;

wherein

said approximate shape information encoded by said approximate shape encoding step and said quantization error encoded by said error encoding step are output.

17. A decoding apparatus for decoding encoded audio and/or video signals in which the approximate shape information of the subsidiary information that is not the direct subject-matter of encoding is encoded in an encoding apparatus and in which a quantization error in encoding the approximate shape information is encoded in a lossless fashion without losing the information, said approximate shape information and the quantization error being input to and decoded by the decoding apparatus, said decoding apparatus comprising:

error decoding means for decoding the quantization error of said subsidiary information;

approximate shape decoding means for decoding the approximate shape information of said subsidiary information; and

synthesizing means for synthesizing the decoded quantization error and the decoded approximate shape information.

18. The decoding apparatus according to claim 17 wherein

said approximate shape information of said subsidiary information is found from linear prediction coefficients of said subsidiary information.

19. The decoding apparatus according to claim 17 wherein

the approximate shape of said subsidiary information is that found from the cepstrum coefficients of said subsidiary information.

20. The decoding apparatus according to claim 17 wherein

the approximate shape of said subsidiary information is that found by calculating the representative value of said subsidiary information with respect to each preset unitary unit.

21. The decoding apparatus according to claim 20 wherein

said representative value is an average value.

22. The decoding apparatus according to claim 20 wherein

said representative value is a maximum value.

23. The decoding apparatus according to claim 17 wherein

in encoding the approximate shape, the approximate shape information is vector-quantized.

24. The decoding apparatus according to claim 23 wherein

from a codebook for encoding the approximate shape, provided from the outset, an index of a codeword which minimizes the quantization error is encoded.

25. The decoding apparatus according to claim 17 wherein

said quantization error is the difference between the subsidiary information and the encoded approximate shape information.

26. The decoding apparatus according to claim 17 wherein

in encoding said quantization error, said quantization error is encoded into codewords of a variable length.

27. The decoding apparatus according to claim 26 wherein

the codewords of a variable length are composed of Huffman code.

28. The decoding apparatus according to claim 17 wherein

said subsidiary information is the normalization information.

29. The decoding apparatus according to claim 17 wherein

said subsidiary information is the quantization step information.

30. The decoding apparatus according to claim 17 further comprising:

separating means for separating the encoded audio and/or video signals into quantization coefficients, quantization step information and normalization information;

decoding means for decoding said quantization coefficients, based on the decoded quantization step information and the decoded normalization information, from one preset encoding unit to another; and

restoration means for synthesizing the decoded quantization coefficients from one encoding unit to another to restore said audio and/or video signals.

31. A decoding method for decoding encoded audio and/or video signals in which the approximate shape information of the subsidiary information that is not the direct subject-matter of encoding is encoded in an encoding apparatus and in which a quantization error in encoding the approximate shape information is encoded in a lossless fashion without losing the information, said approximate shape information and the quantization error being input and decoded, said decoding method comprising:

an error decoding step of decoding the quantization error of said subsidiary information;

an approximate shape decoding step of decoding the approximate shape information of said subsidiary information; and

a synthesizing step of synthesizing the decoded quantization error and the decoded approximate shape information.

32. An encoding program for encoding the main information of audio and/or video signals, which is the direct subject-matter of encoding, and the subsidiary information, which is not the direct subject-matter of encoding, said encoding program comprising:

an approximate shape computing step of computing the approximate shape information of said subsidiary information;

an approximate shape encoding step of quantizing and encoding said approximate shape information;

an error computing step of computing the quantization error; and

an error encoding step of encoding said quantization error in a lossless fashion without losing the information;

wherein

said approximate shape information encoded by said approximate shape encoding step and said quantization error encoded by said error encoding step are output.

33. An decoding program for decoding encoded audio and/or video signals, in which the approximate shape information of the subsidiary information that is not the direct subject-matter of encoding is encoded and in which a quantization error in encoding the approximate shape information is encoded in a lossless fashion without losing the information, said approximate shape information and the quantization error being input and decoded, said decoding program comprising:

an error decoding step of decoding the quantization error of said subsidiary information;

an approximate shape decoding step of decoding the approximate shape information of said subsidiary information; and

a synthesis step of synthesizing the decoded quantization error and the decoded approximate shape information.