Speech and image compressions by DCT, wavelet, and wavelet packet

In this paper, the lossy compression methods of speech and image signals by DCT, WT, and WPT are compared for finding the effective compression method. After lossy compression using the threshold criterion, the compressed speech and image signals were reconstructed. The reconstructed signals are measured by SNR and PSNR. The results have showed that for both speech and image signals, WT and WPT have a better performance than DCT. WPT gives the best PSNR among others except for the computational complexity.