Fast Block Coder Optimized Truncation Reducing Complexity

The patent segments the code-block into multiple stripes and processes them in parallel using multiple thread contexts. Each thread independently codes a portion of the significance information, allowing the complex EBCOT algorithm to be divided into manageable segments that can be executed concurrently, thereby reducing overall computational complexity while maintaining compression efficiency

The patent performs preliminary significance coding before magnitude coding, and pre-organizes the code bits into the final representation format during the significance coding phase. This preliminary organization of data structures and coding sequences eliminates the need for complex post-processing operations, reducing computational complexity while preserving the compression efficiency of the standard EBCOT algorithm

The code-block is divided into multiple stripes that can be processed in parallel. Each stripe is further segmented into processing units that can be handled by different thread contexts simultaneously. This segmentation enables concurrent execution of the EBCOT algorithm across multiple processing units, significantly improving processing speed while maintaining the compression efficiency of the original algorithm

The patent implements continuous processing by eliminating idle periods between coding operations. The significance information is coded continuously across all stripes without interruption, and the code bit organization is performed continuously during the coding process rather than in separate batch operations. This continuous useful action maximizes processing speed while preserving compression efficiency