New Clustering Algorithm ‘TCEVR’ for Vector Quantization using Cosine Transform

Vector Quantization (VQ) plays important role in codebook generation such that the distortion between the original image and the reconstructed image is the minimum. In this paper we present an effective clustering algorithm to generate codebook for vector quantization. In existing algorithm KEVR while splitting the cluster every time new orientation is introduced using error vector sequence. This error vector sequence is binary representation of numbers, so cluster orientation change slowly in every iteration. KEVRW uses Walsh sequence to rotate the error vector. Because of this cluster orientation change rapidly in every iteration. The proposed technique TCEVR (Thepade’s Cosine error vector rotation) is based on KEVR algorithm. The proposed methodology is tested on different training images for various codebook of sizes 64, 128, 256, 512.The obtained results shows that TCEVR gives less MSE as well as less distortion as compared to KEVR, KEVRW.