Block-conditioned context-based arithmetic coding for efficient representation of binary shapes

In this paper, we propose a block-conditioned context-based arithmetic-encoding (BCAE) scheme to attain efficient binary shape coding. By performing simple block detection, we first setup an efficient conditional source model in a block-by-block fashion such that each block acquires extra side information to characterize the block mode. With a re-trained probability table for each mode, simulations show that the coding efficiency of the proposed BCAE algorithms is better than that of the MPEG-4 context-based arithmetic-encoding (CAE) method. To enhance the coding efficiency, the tag tree encoding scheme is also suggested to further condense the extra side information. Confirmed by simulations, the proposed BCAE algorithm with the tag tree block mode encoding method could be a good tool for future binary shape coding.