Image compression using transformed vector quantization

摘要

Vector quantization (VQ) is an important technique in digital image compression. To improve its performance, we would like to speed up the design process and achieve the highest compression ratio possible. To speed up the process, we used a fast Kohonen self-organizing neural network algorithm to achieve big saving in codebook construction time. To obtain better reconstructed images, we propose a new approach called the transformed vector quantization (TVQ), combining the features of transform coding and VQ. We use several data sets to demonstrate the feasibility of this TVQ approach. A comparison of reconstructed image quality is made between the TVQ and VQ. Also, a comparison is made between a TVQ and a standard JPEG approach.