Picture coding method and picture decoding method
View Patent ↗The method includes the following units: a coefficient number detecting unit ( 109 ) for detecting the number of coefficients which has a value other than 0 for each block according to the generated coefficient, a coefficient number storing unit ( 110 ) for storing the number of coefficients detected, a coefficient number coding unit ( 111 ) for selecting a table for variable length coding based on the numbers of coefficients in the coded blocks located on the periphery of a current block to be coded with reference to the selected table for variable length coding so as to perform variable length coding for the number of coefficients.
1. A decoding method for decoding coded data on a block basis, the method comprising:
determining whether a first decoded block located above a target block and a second decoded block located to the left of the target block are located outside a slice including the target block;
predicting a predictive value for a total number of non-zero coefficients included in the target block to be decoded based on a total number of non-zero coefficients included in the first decoded block and a total number of non-zero coefficients included in the second decoded block when both the first decoded block and the second decoded block are inside said slice, and predicting the predictive value based on the total number of non-zero coefficients included in one of the first decoded block and the second decoded block when the other of the first decoded block and the second decoded block is outside said slice, wherein each non-zero coefficient is a transform coefficient having a level value other than “0”;
selecting a variable length code table based on the predictive value; and
decoding the coded data by using the selected variable length code table to obtain an actual value of a total number of non-zero coefficients included in the target block.
2. A picture reconstructing method for reconstructing a picture from coded data obtained by coding a picture on a block basis, the method comprising:
decoding the coded data to obtain a total number of non-zero coefficients included in a first block located above a target block and a total number of non-zero coefficients included in a second block located to the left of the target block, each non-zero coefficient being a transform coefficient having a level value other than “0”;
determining whether the first decoded block and the second decoded block are located inside a slice including the target block;
predicting, according to said determining, a predictive value for a total number of non-zero coefficients included in the target block to be decoded based on the total number of non-zero coefficients included in the first block and the total number of non-zero coefficients included in the second block when both the first block and the second block are inside said slice, and predicting the predictive value based on the total number of non-zero coefficients included in one of the first block and the second block when the other of the first block and the second block is outside said slice;
selecting a variable length code table based on the predictive value;
decoding the coded data to obtain an actual value of a total number of non-zero coefficients included in the target block;
decoding the coded data by using the obtained actual value to obtain a decoded target block; and reconstructing a reconstructed picture by using the decoded target block.