Abstract
Vector Quantization (VQ) is an efficient technique for image compression. A new fast VQ scheme, called DCT-VQ, based on a two-dimensional discrete cosine transform (2-D DCT) is presented. The significant features of training images are extracted by using the 2-D DCT. A codebook is generated by partitioning the training set into a binary tree. Each training vector at a nonterminal node of the binary tree is distributed to one of the two descendants by comparing a single feature associated with that node to a threshold. A modified scheme based on DCT-VQ is proposed. It partitions the training data into a binary tree according to the distribution of training vectors. Experimental results show that the new scheme results in a better codebook