A MS-GS VQ codebook design for wireless image communication using genetic algorithms

An image compression technique is proposed that attempts to achieve both ro bustness to transmission bit errors common to wireless image communication, as well as sufficient visual quality of the reconstructed images, Error ro bustness is achieved by using biorthogonal wavelet subband image coding wit h multistage gain-shape vector quantization (MS-GS VQ) which uses three sta ges of signal decomposition in an attempt to reduce the effect of transmiss ion bit errors by distributing image information among many blocks, Good vi sual quality of the reconstructed images is obtained by applying genetic al gorithms (GA's) to codebook generation to produce reconstruction capabiliti es that are superior to the conventional techniques, The proposed decomposi tion scheme also supports the use of GA's because decomposition reduce the problem size, Some simulations for evaluating the performance of the propos ed coding scheme on both transmission bit errors and distortions of the rec onstructed images are performed. Simulation results show that the proposed MS-GS VQ with good codebooks designed by GA's provides not only better robu stness to transmission bit errors but also higher peak signal-to-noise rati o even under high bit error rate conditions.