Digital image wavelet compression enhancement via particle swarm optimization

Data compression is essentially a technical solution to confront the needs of the storage capacity and data redundancy. As a typical multimedia technique, compression on digital images and videos plays an important role in data compression. Some lossy image compression methods are acceptable for achieving a substantial reduction in bit rate, with minor information loss. From the point of view of compression, wavelet transform has emerged as a dominant technology in the fields of signal and image processing, such as the discrete wavelet transform and wavelet packet decomposition. However, a regular transform is limited by its wavelet bases which downsize by a power of two towards the low frequencies, thus it may not produce best results. Being a stochastic population based evolutionary algorithm, the particle swarm optimization (PSO) simulates social optimization, where the swarm is modeled by particles in multidimensional space together with a position and a velocity. In this case, it has been proposed to enhance the quality of the wavelet based image compression. Objective measures are also presented to evaluate the integration of the wavelet expansion technologies and PSO optimization scheme.