Image Compression Using Neural Networks

Abstract

Digital Communication involves sending and receiving of data bits over long distances using various communication channels. Image files are commonly sent over such Digital channels. Storage and transfer of digital images involves a very large number of bits. This presents a heavy load on the network. Hence Image Compression is commonly required. It is a subdomain of data compression. Existing traditional techniques have both lossy and lossless forms. In this paper we are discussing a Technique for Compressing Images using Artificial Neural Networks. The proposed technique uses gradient descent and genetic algorithm approaches and attempts to overcome the problem of traditional techniques and either assist or replace the traditional techniques for Image Compression.