Certain investigations on VLSI architecture design for two dimensional discrete wavelet transform and entropy coder

Abstract

A raw image, in its original form, includes huge amount of information which requires not only a large amount of memory for its storage but also makes inconvenient transmission over limited channel bandwidth. Image compression removes the data redundancy from the image in either lossy or lossless methods. Lossless image compression can retrieve the original image exactly with low compression and lossy image compression technique compresses the image in a variable amount based on the quality of image required for its use in specific application area. It is processed in a series of steps such as image transform coding, quantization and entropy coding. One of the famous image compression standards, JPEG 2000 adopts Discrete Wavelet Transform (DWT) for transformation of the image from spatial domain to frequency domain. In general, an image contains a high visual information concentrated on low frequencies and poor visual information in its high frequencies. The quantization process can be done after the image transformation to reduce the size of transformed wavelet coefficients. Furthermore, entropy coding is performed using arithmetic encoding method to reduce the redundancy that exists both in the transformed and quantization data path. newline newline newline