CODING METHOD, DECODING METHOD, CODEC METHOD, CODEC SYSTEM AND RELEVANT APPARATUSES

• US 20110093275A1
• Filed: 12/30/2010
• Published: 04/21/2011
• Est. Priority Date: 11/10/2008
• Status: Active Grant

• Current Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Inventors: Li, Haiting
• Granted Patent: US 8,731,947 B2
• Time in Patent Office: Days
• US Class Current: 704/500
• CPC Class Codes: G10L 19/038 Vector quantisation, e.g. TwinVQ audio
• CPC Class Codes: G10L 2019/0005 Multi-stage vector quantisation
• CPC Class Codes: G10L 2019/0007 Codebook element generation
• CPC Class Codes: G10L 2019/0013 Codebook search algorithms
• CPC Class Codes: H03M 7/3082 Vector coding for television signals, see H04N19/94

Abstract

A coding method, a decoding method, a coding-decoding (codec) method, a codec system and relevant apparatuses are disclosed. The coding method includes: obtaining an amplitude vector and a length vector corresponding to a vector to be coded; sorting elements of the amplitude vector and elements of the length vector; and obtaining a position index value according to the sorted amplitude vector and the sorted length vector. A decoding method, a codec system, and relevant apparatuses are also provided.

Claims

1. A coding method, comprising:

obtaining an amplitude vector and a length vector corresponding to a vector to be coded;

sorting elements of the amplitude vector and elements of the length vector; and

obtaining a position index value according to the sorted amplitude vector and the sorted length vector.

2. The method according to claim 1, wherein:

the sorting of the elements of the amplitude vector and the elements of the length vector comprises;

sorting the elements of the amplitude vector and the elements of the length vector according to a preset sorting rule; and

the preset sorting rule comprises sorting the elements from large values to small values or sorting the elements from small values to large values.

3. The method according to claim 1, wherein:

the obtaining of the position index value according to the sorted amplitude vector and the sorted length vector comprises;

performing permutation coding for positions of elements in the vector to be coded in best order of element removal according to the sorted amplitude vector and the sorted length vector, and obtaining the position index value.

4. The method according to claim 1, further comprising:

searching a preset codebook index value offset table to obtain a codebook index value offset corresponding to the vector to be coded; and

obtaining a lattice codebook index value corresponding to the vector to be coded according to the codebook index value offset and the position index value.

5. The method according to claim 3, further comprising:

searching a preset codebook index value offset table to obtain a codebook index value offset corresponding to the vector to be coded; and

obtaining a lattice codebook index value corresponding to the vector to be coded according to the codebook index value offset and the position index value.

6. The method according to claim 1, wherein the obtaining of the amplitude vector and the length vector corresponding to the vector to be coded comprises:

finding a leader vector of the vector to be coded; and

obtaining the corresponding amplitude vector and length vector according to the leader vector, wherein;

the amplitude vector indicates elements of different values in the leader vector, and the length vector indicates how many times each element in the amplitude vector appears in the leader vector.

7. The method according to claim 2, wherein the obtaining of the amplitude vector and the length vector corresponding to the vector to be coded comprises:

finding a leader vector of the vector to be coded; and

obtaining the corresponding amplitude vector and length vector according to the leader vector, wherein;

the amplitude vector indicates elements of different values in the leader vector, and the length vector indicates how many times each element in the amplitude vector appears in the leader vector.

8. The method according to claim 3, wherein the obtaining of the amplitude vector and the length vector corresponding to the vector to be coded comprises:

finding a leader vector of the vector to be coded; and

obtaining the corresponding amplitude vector and length vector according to the leader vector, wherein;

the amplitude vector indicates elements of different values in the leader vector, and the length vector indicates how many times each element in the amplitude vector appears in the leader vector.

9. The method according to claim 2, wherein:

the sorting of the amplitude vector and the length vector according to the preset sorting rule comprises;

rearranging the elements in the amplitude vector, and adjusting order of the elements in the length vector according to order of the rearranged elements in the amplitude vector; and

rearranging the elements in the length vector whose element order is adjusted, and adjusting the order of the elements in the amplitude vector according to the order of the rearranged elements in the length vector.

10. The method according to claim 2, wherein:

the sorting of the amplitude vector and the length vector according to the preset sorting rule comprises;

rearranging the elements in the length vector, and adjusting order of the elements in the amplitude vector according to order of the rearranged elements in the length vector; and

if several elements are the same in the rearranged length vector, adjusting order of the elements in the amplitude vector corresponding to the elements which are the same in the length vector according to the preset rule.

11. The method according to claim 3, wherein:

the performing of permutation coding for the positions of elements in the vector to be coded in the best order of element removal according to the sorted amplitude vector and the sorted length vector and obtaining of the position index value comprises;

dividing the vector to be coded into layers, wherein number of the layers is equal to number of dimensions of the amplitude vector;

using the vector to be coded as a top-layer vector;

removing the elements in the sorted amplitude vector corresponding to the elements arranged from large values to small values in the sorted length vector to obtain a new vector, and using the new vector as a vector on a layer next to a top layer, wherein the removing starts from the top-layer vector;

removing the elements on lower layers until all layers are finished; and

calculating the position index value according to results of the layering and number of dimensions of the new vector on each layer.

16. The method according to claim 11, wherein:

the sorting of the elements of the amplitude vector and the length vector comprises;

sorting the elements of the amplitude vector and the elements of the length vector according to a preset sorting rule;

wherein the preset sorting rule is to sort the elements from large values to small values or to sort the elements from small values to large values.

12. A decoding method, comprising:

receiving a lattice codebook index value sent by a coder;

obtaining an amplitude vector and a length vector;

sorting elements of the amplitude vector and elements of the length vector; and

performing decoding according to the sorted amplitude vector and the sorted length vector to obtain a vector.

13. The method according to claim 12, wherein:

the obtaining of the amplitude vector and the length vector comprises;

searching a preset codebook index value offset table to obtain a maximum codebook index value offset less than the lattice codebook index value; and

obtaining element values incorporated in the vector, the amplitude vector, and the length vector according to the found codebook index value offset.

14. The method according to claim 13, further comprising:

subtracting the codebook index value offset from the lattice codebook index value to obtain a new index value;

the performing of decoding according to the sorted amplitude vector and the sorted length vector to obtain the vector comprises;

performing decoding according to the new index value, the sorted amplitude vector, and the sorted length vector to obtain the vector.

15. The method according to claim 14, wherein:

after subtracting the codebook index value offset from the lattice codebook index value to obtain the new index value, the method comprises;

separating a sign index value and a position index value from the new index value; and

decoding out a sign vector of nonzero elements according to the sign index value;

the performing of decoding according to the new index value, the sorted amplitude vector, and the sorted length vector to obtain the vector comprises;

performing decoding according to the position index value, the sorted amplitude vector, and the sorted length vector to obtain a signless vector; and

calculating out a vector according to the sign vector of nonzero elements and the signless vector.

17. A coding apparatus, comprising:

a searching unit, configured to search a preset codebook index value offset table to obtain a codebook index value offset corresponding to a vector to be coded;

an obtaining unit, configured to obtain an amplitude vector and a length vector corresponding to the vector to be coded;

a sorting unit, configured to sort elements of the amplitude vector and elements of the length vector obtained by the obtaining unit;

a permutation coding unit, configured to perform permutation coding for positions of elements in the vector to be coded in best order of element removal according to the amplitude vector and the length vector sorted by the sorting unit, and obtain a position index value; and

an executing unit, configured to calculate a lattice codebook index value corresponding to a final vector to be coded according to the codebook index value offset found by the searching unit and the position index value obtained by the permutation coding unit, and send the lattice codebook index value to a decoding apparatus.

18. The coding apparatus according to claim 17, further comprising:

a separating unit, configured to separate a sign vector from the vector to be coded; and

a sign coding unit, configured to encode the separated sign vector separately to obtain a sign index value;

wherein the obtaining unit is configured to obtain an amplitude vector and a length vector corresponding to a signless vector to be coded;

a permutation coding unit, configured to perform permutation coding sequentially for positions of elements in the signless vector to be coded to obtain a position index value; and

an executing unit, configured to calculate a lattice codebook index value corresponding to a final vector to be coded according to the codebook index value offset, the position index value, and the sign index value.

19. A decoding apparatus, comprising:

a receiving unit, configured to receive a lattice codebook index value sent by a coding apparatus;

a searching unit, configured to search a preset codebook index value offset table to obtain a maximum codebook index value offset less than the received lattice codebook index value, and obtain element values incorporated in a vector, an amplitude vector, and a length vector according to the found codebook index value offset;

a generating unit, configured to subtract the codebook index value offset from the lattice codebook index value to obtain a new index value;

a sorting unit, configured to sort elements of the amplitude vector and elements of the length vector obtained by the searching unit; and

a decoding unit, configured to perform decoding according to the new index value obtained by the generating unit as well as the amplitude vector and the length vector sorted by the sorting unit to obtain a vector.

20. The decoding apparatus according to claim 19, further comprising:

an index value separating unit, configured to separate a sign index value and a position index value from the new index value generated by the generating unit;

wherein the decoding unit is configured to decode a sorting result of the sorting unit according to the position index value to obtain a signless vector, perform decoding according to the sign index value to obtain a sign vector of nonzero elements, and then calculate out a vector according to the obtained sign vector of nonzero elements and the signless vector.