Vol. 2, Issue 5 (2015)
DWT based set partitioning in hierarchical trees algorithm for image compression
Author(s): Nandini Dhole, Kunal Deshmukh
Abstract: Embedded zero-tree wavelet (EZW) coding, proposed by J. M. Shapiro, is a very computationally simple and effective technique for image compression. Now we express an alternative description of the principles, so that the purpose for its better performance can be well understood. These principles are ordered bit plane transmission, an image wavelet transform self-similarity across different scales and a set partitioning sorting algorithm as a partial ordering by magnitude. Also we present an optional and efficient implementation based on set partitioning in hierarchical trees (SPIHT), image compression algorithm which is very effective, less complexity and generates an embedded compressed bit-stream that can be effectively decoded at several data rates. The results of image coding, obtained from reconstructed images by the decoding algorithm and actual file sizes, are comparable to or surpass previous results obtained through much more computationally complex and sophisticated methods. In addition, this new coding and decoding techniques are very fast.