Fast Codebook Generation Using Pattern Based Masking Algorithm for Image Compression

Vector Quantization (VQ) is a classical block coding technique used for image compression which achieves high compression using simple encoding and decoding process. Codebook generation is an important factor in VQ design, which directly influences computational cost and the quality of the reconstructed image. Linde-Buzo-Gray (LBG) is considered as a state of art technique, which uses k-mean clustering algorithm for codebook design. Various optimization techniques are applied for searching the optimal codebook, such as Bat Algorithm (BA), Particle swarm optimization (PSO), and Firefly Algorithm (FA). These algorithm suffers mainly with high time consumption due to unavailability of the optimal solution in search space. This research proposes a novel approach, where peak values of the histogram are applied to predefined pattern masks to predict the image patterns for codebook design. From the experimental results, it is indicated that when compared with other algorithms, the proposed pattern based masking (PBM) algorithm requires fewer iterations and converges at a faster speed, particularly at the bitrates >= 0.375 without compromising on peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM).