Color Images Of Three-dimensional Compression Technology And Adaptive Extraction Of Roi

In this paper a method based on discrete wavelet transform and discrete cosine transform is presented to compress color image. Color image can be thought to have three spectra. A three-dimensional transform (discrete wavelet transform in the spatial domain and discrete cosine transform in the spectral domain) is taking and the transformed coefficients using three-dimensional algorithm are encoded. The experiment shows that DCT on color space has better performance than YUV and RGB. SPECK is an efficient embedded encoding algorithm and has been improved so as to fit it for color image encoding. Comparing the improved SPECK with 3-D SPIHT is given by computer simulation. The experimental results suggest that SPECK has higher encoding speed and lower memory requirement. Furthermore, a simple method is presented to detect the region of interest in an image adaptively. The experimental results show that its effect is good. Combining this with ROI technique make the visual quality of uncompressed image improving.