Edge Preserving Image Compression Technique Using Feed Forward Neural Network

Abstract

With the growth of multimedia and internet, compression techniques have become the thrust area in the fields of computers. Popularity of multimedia has led to the integration of various types of computer data. Multimedia combines many data types like text, graphics, still images, animation, audio and video. Image compression is a process of efficiently coding digital image to reduce the number of bits required in representing image. Its purpose is to reduce the storage space and transmission cost while maintaining good quality. Many different image compression techniques currently exist for the compression of different types of images. In the present research work back propagation neural network training algorithm has been used. The neural network model has been trained and tested for the different types of images. Back propagation neural network algorithm helps to increase the performance of the system and to decrease the convergence time for the training of the neural network. The aim of this work is to develop an edge preserving image compressing technique using one hidden layer feed forward neural network of which the neurons are determined adaptively .The processed image block is fed as a single input pattern while single output pattern has been constructed from the original image unlike other neural network based technique where multiple image blocks are fed to train the network. The initialization of weights between the lone hidden layer by transforming pixel coordinates of the input pattern block into its equivalent one dimensional representation. The initialization process exhibit better rate convergence of the back propagation training algorithm as compare to the randomization of initial weight. The proposed scheme has been demonstrated through several experiments including lena, girl, cameraman and very promising results in compression as well as in reconstructed image over convectional neural network based technique are obtained.