Improve Wavelet Threshold Selection Based on PSO Algorithm and Successive –Approximation Method in Image Compression Applications

The growing rate in the applications that use the image compression such as the internet, multimedia, satellite imaging, medical imaging etc., make this subject under heuristics of researchers. So, this field of image compression has been continuous to grow rapidly at the last years. From the compression point of view, the wavelet transform plays an important role and a dominant technology in the fields of signal and image processing. It helps us precisely decompose the signal into low and high frequency components making the thresholding process for redundant data done in an efficient way. Also, the threshold value selection method performs a major part in data compression system. Therefore, in this paper, two approaches have been proposed. The first one depends on implementation the discrete wavelet transform (DWT) with Particle Swarm Optimization (PSO) algorithm to get a good selection for the threshold. The second is a new successive-approximation threshold (SA-thresh) selection method used to produce the global threshold value for the wavelet detail coefficients of the image. These proposed methods give more improvement in compression ratio (CR) for the image as well as enhancement in the peak signal to noise ratio (PSNR). The implementation of this work has been done by using MATLAB 2011a