A Hybrid technique for near-lossless image compression: rounding the
intensity, dividing, and lempel-ziv-welch (RIFD-LZW)
Abstract
Abstract— To provide near-lossless compression for
both color and grayscale images, this study offers RIFD-LZW, a novel
hybrid image compression algorithm that combines the Rounding the
Intensity and Dividing (RIFD) technique with Lempel-Ziv-Welch (LZW)
encoding. The methodology achieves a higher compression ratio than
either technique alone as it uses RIFD to reduce image redundancy and
LZW to efficiently encode data. When applied to the Kodak dataset,
experimental results show that RIFD-LZW decreases file size by 73.03%
and improves compression ratio by 9.7% when compared to RIFD-Huffman.
Additionally, it attains a compression ratio for grayscale images that
is 29.6% greater. With average Mean Squared Error (MSE) values of 2.88,
2.93, and 2.78 for the EPFL, Kodak, and Waterloo datasets, respectively,
the proposed algorithm keeps up its outstanding visual quality. The tiny
distortion that is introduced matches the MSE of RIFD and is
undetectable to the human visual system. These results validate that
RIFD-LZW is a strong contender for both lossy and near-lossless
compression applications since it offers effective image compression
with good image fidelity preservation.