An image compression model via adaptive vector quantization: hybrid optimization algorithm

Abstract

ABSTRACT The Vector Quantization technique, employed in this paper is based on an upgraded Linde–Buzo–Gray (LBG) with a DCT method to store the image data. The code-books are optimized using a unique approach called hybrid optimization. Here, the database image is divided into a series of blocks called vectors. Then, a suitable codeword is chosen as the closest approximation of the input vector. The encoder sends the compressed stream, i.e. indices of these vectors to the decoder. The decoder subsequently decodes the index to obtain the compressed vector. Moreover, the Flower Pollination Adapted Rider Optimization (FPARA) model is used to improve the codebook for providing best image compression result. The optimization of code books is done to lower the summation of compression ratio and the error difference between the original and decompressed images. Therefore, the adopted approach is computed to the traditional schemes, and the simulation outcomes are validated.