A Proposed Image Compression Technique Based on DWT and Predictive Techniques

The idea behind the suggested developed image compression system depends on the application of DWT on error prediction values (Residual Image) instead of the original image. The error prediction values have simpler values compared to the original image, to reduce the coefficient values of all the wavelet bands. As a result, the detail bands can overcome for the first decomposition level and all / some of detail bands for the second decomposition level. In addition, the effect of quantization process is less. The system has been applied on the natural and gray-scale images. The results obtained was similar for both image types. Seven types of different wavelet for four families were compared: {dbl, db3, db5, syml, coifl, bior 2.2, bi02.4.}. The db3 wavelet produced the best (PSNR), which nearest from Coifl wavelet.