Image Compression Using Wavelet Method & SPIHT Algorithm

A variety of orthogonal and biorthogonal filters have been used for signal analysis and compression in wavelet based image coding. The selection of wavelet filter plays the most important part for a particular image. We are trying to find out best wavelet filter at different input rates, different level of decomposition and by eliminating horizontal (H), vertical (V) or diagonal (D) components permanently at higher compression ratio (CR) but with tolerable amount of information loss. The optimized filter will be selected on the basis of peak signal-to-noise ratio (PSNR) value and psycho visual results. Here the wavelet filters to be implemented are sym5, bior4.4, Haar, db5, sym8 and coif1. The input rates can be varied between the interval (0,1) and level of decomposition can be either 1 or 2. The proposed method is a lossy compression technique i.e. the redundancy is permanently removed and can't be achieved back on reconstruction.