Image Compression Based on an Improved SPIHT Algorithm

摘要

Necessity of image compression is described. Wavelet transform is adopted for image compression. Extensive simulations are performed to justify the choice of the D 9/7 wavelet. SPIHT (Set Partition in Hierarchical Trees) is chosen for image coding and then improved to address its defects. Desirable PSNR (Power Signal-to-Noise Ratio) is achieved and the image is reconstructed clearly in the MATLAB simulations. Analysis of experimental data indicates that the proposed method achieves considerable improvement in PSNR.