Wavelet and learning based image compression systems

Abstract

Image compression is a critical element in storage, retrieval and transmission applications. The list of traditional approaches to image compression has already been expanded by wavelet and learning based systems. Here, we report a few techniques which are based on discrete wavelet transform (DWT), Artificial Neural Network (ANN) in feedforward and unsupervised form. The experiments are repeated with images mixed with salt and pepper noise and the outcomes are compared. The quality of the image compression systems is determined by finding the mean square error (MSE), Peak Signal to Noise Ratio (PSNR) and Compression Ratio (CR).