A novel Vector Quantization (VQ) technique for encoding the Bi-orthogonal wavelet decomposed image using hybrid Adaptive Differential Evolution (ADE) and a Pattern Search optimization algorithm (hADEPS) is proposed. ADE is a modified version of Differential Evolution (DE) in which mutation operation is made adaptive based on the ascending/descending objective function or fitness value and tested on twelve numerical benchmark functions and the results are compared and proved better than Genetic Algorithm (GA), ordinary DE and FA. ADE is a global optimizer which explore the global search space and PS is local optimizer which exploit a local search space, so ADE is hybridized with PS. In the proposed VQ, in a codebook of codewords, 62.5% of codewords are assigned and optimized for the approximation coefficients and the remaining 37.5% are equally assigned to horizontal, vertical and diagonal coefficients. The superiority of proposed hybrid Adaptive Differential Evolution and Pattern Search (hADE-PS) optimized vector quantization over DE is demonstrated. The proposed technique is compared with DE based VQ and ADE based quantization and with standard LBG algorithm. Results show higher Peak Signal-to-Noise Ratio (PSNR) and Structural Similiraty Index Measure (SSIM) indicating better reconstruction.
Linde–Buzo–Gray (LBG), a traditional method of vector quantization (VQ) generates a local optimal codebook which results in lower PSNR value. The performance of vector quantization (VQ) depends on the appropriate codebook, so researchers proposed optimization techniques for global codebook generation. Particle swarm optimization (PSO) and Firefly algorithm (FA) generate an efficient codebook, but undergoes instability in convergence when particle velocity is high and non-availability of brighter fireflies in the search space respectively. In this paper, we propose a new algorithm called BA-LBG which uses Bat Algorithm on initial solution of LBG. It produces an efficient codebook with less computational time and results very good PSNR due to its automatic zooming feature using adjustable pulse emission rate and loudness of bats. From the results, we observed that BA-LBG has high PSNR compared to LBG, PSO-LBG, Quantum PSO-LBG, HBMO-LBG and FA-LBG, and its average convergence speed is 1.841 times faster than HBMO-LBG and FA-LBG but no significance difference with PSO.
Linde-Buzo-Gray (LBG) Vector Quantization (VQ), technically generates local codebook after many runs on different sets of training images for image compression. The key role of VQ is to generate global codebook. In this paper, we present comparative performance analysis of different optimization techniques. Firefly and Cuckoo search generate a near global codebook, but undergoes problem when non-availability of brighter fireflies and convergence time is very high respectively. Hybrid Cuckoo Search (HCS) algorithm was developed and tested on four benchmark functions, that optimizes the LBG codebook with less convergence rate by taking McCullochs algorithm based levy flight and variant of searching parameters. Practically, we observed that Bat algorithm (BA) peak signal to noise ratio is better than LBG, FA, CS and HCS in between 8 to 256 codebook sizes. The convergence time of BA is 2.4452, 2.734 and 1.5126 times faster than HCS, CS and FA respectively.