Efficient Lossy Image Compression using Vector Quantization (ELIC-VQ)

Compression is the technique for effective utilization of space in servers as well as in personal computers. Most significantly, being the multimedia compression. In this paper, the focus is on Image Compression method. Image compression method has two types: lossy and lossless compression. Vector quantization is an effective way of lossy compression technique. The important tasks in vector quantization are codebook generation and searching. LBG algorithm is a prominent standard for vector quantization. The major drawback with LBG compression is complexity in computation, which is directly proportional to size of the codebook and number of pixels in image. Another drawback of LBG is global codebook generation which is time consuming and standardizing this codebook is not possible. A novel method is proposed in this paper to address these issues. The proposed method is Efficient Lossy Image Compression using Vector Quantization (ELIC-VQ). It generates global codebook and uses centroid based approach to remove local problem of optimization.  A centroid based compression reduces the operation of the comparison with the codebook and helps to improve the performance. At the time of decompression of the image, the codebook comparison is dependent on the index similar to LBG. The experimental results show that ELIC-VQ approach reduces the computational complexity, increases compression percentage and speed up the vector quantization process. The reconstructed image has reduced distortion significantly than using LBG.