Convergence Analysis of Codebook Generation Techniques for Vector Quantization using K-Means Clustering Technique

Vector Quantization (VQ) is one of the lossy image compression techniques. VQ comprises of three different phases: Codebook Generation, Image Encoding and Image Decoding. The performance of VQ is mainly based on the codebook generation phase. In this paper, five different codebook generation techniques namely the Simple Codebook Generation (SCG), Ordered Codebook Generation (OCG), Codebook Generation by Sorting the Sum of Sib Vectors (CBSSSV), Codebook Generation with Edge Features (CBEF) and Codebook Generation with Cluster Density (CBCD) for Vector Quantization have been discussed and their performance in terms of number of iterations required to converge with respect to Peak Signal to Noise Ratio (PSNR) is compared when kMeans Clustering technique is used to optimize the initial codebook that is created by any of the above techniques. Of these discussed techniques, the CBEF technique performs better. General Terms Vector Quantization, Image Compression