Fast Efficient Transforms for Contours Extraction and Image Compression using Zonal Sampling Methods

Abstract

People who use digital image processing techniques usually deal with a huge amount of data. Because storing image data for future needs requires a lot of storage space. In a similar way, sending image data at a reasonable time requires a high-capacity channel. To minimize these needs, it uses a technique called data compression. In this paper we have implemented the digital image compression process using fast walsh transform (FWT) and discrete cosine transform (DCT). The regions selection of compression is done by using of low-pass filter (LPF); while the extracted contours is performed using high-pass filter (HPF) through Matlab programming language. In this work the nine different shapes of zonal sampling in spectral domain will be tested. The comparison between FWT and DCT using different shapes of zonal sampling is implemented in terms of peak signal to noise ratio (PSNR), compression ratio (CR), and bit per pixel (bpp). Experimental results and analysis for image compression using LPF show that DCT gives higher PSNR (45 decibels) with CR=89.0144% than FWT (36 decibels) with CR=79.6921%; while the FWT introduce better quality than that by using DCT for contour extraction using HPF.