Image processing device and image processing method for image encoding based on non-square quantization matrix

Abstract

There is provided an image processing device including a decoding section that decodes an encoded stream and generates quantized transform coefficient data, and an inverse quantization section that, taking transform coefficient data as transform units to be used during inverse orthogonal transform, inversely quantizes the quantized transform coefficient data decoded by the decoding section, such that in a case where a non-square transform unit is selected, the inverse quantization section uses a non-square quantization matrix, corresponding to a non-square transform unit, that is generated from a square quantization matrix corresponding to a square transform unit.

What is claimed is: 
     
         1 . An image processing device, comprising:
 a processor configured to:
 orthogonally transform image data based on a non-square transform unit; 
 generate transform coefficient data based on the orthogonally transformed image data; 
 set a non-square quantization matrix based on a copy of one of row elements at an interval or column elements at the interval of a square quantization matrix; 
 quantize the transform coefficient data based on the non-square quantization matrix; and 
 encode the quantized transform coefficient data to generate a bit stream including the square quantization matrix. 
   
     
     
         2 . The image processing device according to  claim 1 , wherein a size of a long edge of the non-square quantization matrix is equal to a size of an edge of the square quantization matrix. 
     
     
         3 . The image processing device according to  claim 1 , wherein one of the row elements or the column elements of the square quantization matrix are based on a position associated with the square quantization matrix. 
     
     
         4 . The image processing device according to  claim 1 , wherein
 the interval of one of the row elements and the column elements to be copied from the square quantization matrix is determined according to a ratio of a size of a short edge of the non-square quantization matrix versus a size of an edge of the square quantization matrix,   the ratio is 1:4, and   the interval is one of four rows or four columns.   
     
     
         5 . The image processing device according to  claim 1 , wherein
 a size of the square quantization matrix is 4×4, and   a size of the non-square quantization matrix is one of 1×4 or 4×1.   
     
     
         6 . The image processing device according to  claim 1 ,
 wherein a size of the square quantization matrix is 8×8, and   wherein a size of the non-square quantization matrix is one of 2×8 or 8×2.   
     
     
         7 . The image processing device according to  claim 1 ,
 wherein a size of the square quantization matrix is 16×16, and   wherein a size of the non-square quantization matrix is one of 4×16 or 16×4.   
     
     
         8 . The image processing device according to  claim 1 ,
 wherein a size of the square quantization matrix is 32×32, and   wherein a size of the non-square quantization matrix is one of 8×32 or 32×8.   
     
     
         9 . A method, comprising:
 orthogonally transforming, by a processor, image data based on a non-square transform unit;   generating, by the processor, transform coefficient data based on the orthogonally transformed image data;   setting, by the processor, a non-square quantization matrix based on a copy of one of row elements at an interval or column elements at the interval of a square quantization matrix;   quantizing, by the processor, the transform coefficient data based on the non-square quantization matrix; and   encoding, by the processor, the quantized transform coefficient data to generate a bit stream including the square quantization matrix.   
     
     
         10 . The method according to  claim 9 , wherein a size of a long edge of the non-square quantization matrix is equal to a size of an edge of the square quantization matrix. 
     
     
         11 . The method according to  claim 9 , wherein one of the row elements or the column elements of the square quantization matrix are based on a position associated with the square quantization matrix. 
     
     
         12 . The method according to  claim 9 , wherein
 the interval of one of the row elements and the column elements to be copied from the square quantization matrix is determined according to a ratio of a size of a short edge of the non-square quantization matrix versus a size of an edge of the square quantization matrix,   the ratio is 1:4, and   the interval is one of four rows or four columns.   
     
     
         13 . The method according to  claim 9 ,
 wherein a size of the square quantization matrix is 4×4, and   wherein a size of the non-square quantization matrix is one of 1×4 or 4×1.   
     
     
         14 . The method according to  claim 9 ,
 wherein a size of the square quantization matrix is 8×8, and   wherein a size of the non-square quantization matrix is one of 2×8 or 8×2.   
     
     
         15 . The method according to  claim 9 ,
 wherein a size of the square quantization matrix is 16×16, and   wherein a size of the non-square quantization matrix is one of 4×16 or 16×4.   
     
     
         16 . The method according to  claim 9 , wherein
 a size of the square quantization matrix is 32×32, and   a size of the non-square quantization matrix is one of 8×32 or 32×8.