Method and system of high throughput arithmetic entropy coding for video coding

Abstract

A method, system, and articles of high throughput arithmetic entropy coding for video coding uses a non-framewidth raster order or non-raster order to form spatial neighbor probability contexts for entropy coding.

What is claimed is: 
     
         1 . A method of video coding comprising:
 obtaining a plurality of frames of a video sequence, wherein individual frames are formed of data from rows of pixels and are divided into multiple entropy blocks;   within a same entropy block having multiple consecutive pixel rows of a frame without any other entropy block partition within the same entropy block, determining a spatial neighbor probability context to be used to arithmetically entropy code data of a current pixel location within the same entropy block and comprising entropy coding spatial neighbor pixels within the same entropy block in a non-raster order within the same entropy block to entropy code the data of the current pixel location before entropy coding substantially all or all of the pixel locations in a part of a row of pixels within the same entropy block and above and adjacent to the row of pixels having the current pixel location; and   using the spatial neighbor probability context to perform arithmetic entropy encoding or decoding of the data of the current pixel location.   
     
     
         2 . The method of  claim 1  wherein partition borders of the same entropy block are also partition borders of a prediction block, a transform block, or both. 
     
     
         3 . The method of  claim 1  wherein no block partition borders divide pixel rows within the same entropy block that limit which pixels within the same entropy block cannot be used as neighbor pixels for the current pixel. 
     
     
         4 . The method of  claim 1  comprising performing the entropy coding of multiple current pixel locations in parallel on multiple different rows within the same entropy block. 
     
     
         5 . The method of  claim 4  comprising entropy coding multiple current pixel locations simultaneously at different locations along the same frame row at multiple entropy blocks including different parts of at least some of the same pixel rows. 
     
     
         6 . The method of  claim 1  comprising performing the non-raster order to entropy code within entropy blocks on part of the frame and a raster order to entropy code on another part of the same frame. 
     
     
         7 . The method of  claim 1  comprising entropy coding current pixel locations forming an edge of an entropy block comprising using fewer spatial neighbor probability blocks within the same entropy block than current pixel locations within the entropy block and away from the edge. 
     
     
         8 . The method of  claim 1  wherein the frame has multiple ones of the same entropy block, and wherein the entropy blocks are formed with a shape, size, position, or combination of these regardless of the shape, size, and position of other blocks used for non-entropy tasks while encoding or decoding pixel data of the at least one frame. 
     
     
         9 . The method of  claim 1  wherein the entropy blocks are the same shape, position and size as non-entropy blocks used to encode or decode pixel data of the at least one frame. 
     
     
         10 . A computer-implemented system of video coding comprising:
 at least one memory to store a plurality of frames of a video sequence, wherein individual frames are formed of data from rows of pixels and are divided into entropy blocks;   at least one processor communicatively connected to the at least one memory and arranged to operate by:   within a same entropy block having multiple consecutive pixel rows of a frame without any other entropy block partition within the same entropy block, determining a spatial neighbor probability context to be used to arithmetically entropy code data of a current pixel location within the same entropy block and comprising entropy coding spatial neighbor pixels within the same entropy block in a non-raster order within the same entropy block to entropy code the data of the current pixel location before entropy coding substantially all or all of the pixel locations in a part of a row of pixels within the same entropy block and above and adjacent to the row of pixels having the current pixel location; and
 using the spatial neighbor probability context to perform arithmetic entropy encoding or decoding of the data of the current pixel location. 
   
     
     
         11 . The system of  claim 10  wherein partition borders of the same entropy block are also partition borders of a prediction block, a transform block, or both. 
     
     
         12 . The system of  claim 10  wherein no block partition borders divide pixel rows within the same entropy block that limit which pixels within the same entropy block can be used as neighbor pixels for the current pixel. 
     
     
         13 . The system of  claim 10  wherein the at least one processor is to operate by performing the entropy coding of multiple current pixel locations in parallel on multiple different rows within the same entropy block. 
     
     
         14 . The system of  claim 10  wherein the at least one processor is to operate by performing the non-raster order to entropy code within the entropy blocks on a part of the frame and raster-order entropy coding on another part of the same frame. 
     
     
         15 . The system of  claim 10  wherein the frame has multiple entropy blocks, and wherein the individual entropy blocks are limited to the use of the spatial neighbor probability data within its own entropy block to perform the entropy coding. 
     
     
         16 . The system of  claim 10  wherein the frame has multiple entropy blocks, and the entropy blocks are formed with a shape, size, position, or combination of these regardless of the shape, size, and position of other blocks used for non-entropy tasks while encoding or decoding pixel data of the at least one frame. 
     
     
         17 . The system of  claim 10  wherein the frame has multiple entropy blocks, and wherein the at least one processor is to operate by performing a block-level raster order entropy coding of rows of pixels within at least some of the individual entropy blocks. 
     
     
         18 . The system of  claim 10  wherein the smallest entropy block matches the smallest prediction block size. 
     
     
         19 . At least one non-transitory article having a computer-readable media comprising instructions that cause a computing device to operate by:
 obtaining a plurality of frames of a video sequence, wherein individual frames are formed of data from rows of pixels and are divided into multiple entropy blocks;   within a same entropy block having multiple consecutive pixel rows of a frame without any other entropy block partition within the same entropy block, determining a spatial neighbor probability context to be used to arithmetically entropy code data of a current pixel location within the same entropy block and comprising entropy coding spatial neighbor pixels within the same entropy block in a non-raster order within the same entropy block to entropy code the data of the current pixel location before entropy coding substantially all or all of the pixel locations in a part of a row of pixels within the same entropy block and above and adjacent to the row of pixels having the current pixel location; and   using the spatial neighbor probability context to perform arithmetic entropy encoding or decoding of the data of the current pixel location.   
     
     
         20 . The article of  claim 19  wherein the entropy encoding uses three or four neighbor pixels to provide the spatial neighbor probability content including two or three neighbor pixels in the row of pixels above the and adjacent the row of pixels having the current pixel location. 
     
     
         21 . The article of  claim 19  wherein partition borders of the same entropy block are also partition borders of a prediction block, a transform block, or both. 
     
     
         22 . The article of  claim 19  wherein no block partition borders divide pixel rows within the same entropy block that limit which pixels within the same entropy block can be used as neighbor pixels for the current pixel. 
     
     
         23 . The article of  claim 19  comprising performing the entropy coding of multiple current pixel locations in parallel on multiple different rows within the same entropy block. 
     
     
         24 . The article of  claim 23  comprising entropy coding multiple current pixel locations simultaneously at different locations along the same frame row at multiple entropy blocks including different parts of at least some of the same pixel rows. 
     
     
         25 . The article of  claim 19  comprising performing the non-raster order to entropy code within entropy blocks on part of the frame and a raster order to entropy code on another part of the same frame.