Development of new algorithm for efficient medical image compression

Abstract

The amount of medical images generated in digital form increases everyday in this modern era. In order to store and transmit them effectively with required resolution in medical field, an efficient compression algorithm is needed. This thesis is concerned with the application of compression for medical image coding. This research work develops an efficient search pattern named Octagon with Diagonal Cross Hair Search (ODCHS) for effective compression by reducing statistical and special redundancies of Integer Wavelet Transform (IWT) coefficients. The main contribution is the development of ODCHS algorithm to search significant pixels for a particular threshold. For each threshold, it processes two passes; sorting and refinement. The significant pixels are encoded only in the former operation and in the later stage they are refined. It is also designed for lossless and lossy compression so that the progressive transmission is possible like in Set Partitioning In Hierarchical Trees (SPIHT) and Set Partitioned Embedded Block Coder (SPECK). newline