Huffman Based LZW Lossless Image Compression Using Retinex Algorithm

Abstract

This project presents an efficient approach to lossless image compression by integrating the Retinex algorithm with Huffman and LZW encoding techniques. The Retinex algorithm enhances image contrast and illumination consistency, which improves compressibility without sacrificing quality. Following enhancement, the image is encoded using a hybrid method combining LZW (Lempel Ziv Welch) compression and Huffman coding to achieve effective redundancy reduction. LZW efficiently identifies repeating patterns, while Huffman coding further optimizes the bitstream based on symbol frequencies. This dual stage compression ensures high fidelity of the reconstructed image while significantly reducing file size. The proposed method is evaluated using standard image quality metrics such as PSNR and Compression Ratio, demonstrating superior performance in preserving image quality with optimal compression efficiency. This approach is suitable for applications where both image integrity and storage optimization are critical.