Memory efficient set partitioning in hierarchical tree for color image coding

A memory efficient version of color SPIHT with enhanced features is presented. The proposed coder, termed color memory efficient SPIHT (CMESH) extends the use of set partitioning in hierarchical trees to color images. The exciting feature of the algorithm is that it uses a single re-usable list instead of three continuously growing linked lists as in conventional CSPIHT. The list is re-initialized accordingly at the beginning of each coding pass and is exhausted within the pass itself. Besides, these passes are independent of each other, and simple linking between chrominance and luminance allows offsets for chrominance and hence comes later in the initialization of the list. The performance of the proposed coder is measured in terms of coding efficiency and maximum memory requirement due to list entries in each bit-plane. The performance is compared with that of CSPIHT. It is observed that the proposed algorithm results in a saving of 50-70% memory while having comparable coding efficiency.