Near lossless image compression using parallel fractal texture identification

Publication date: April 2020Source: Biomedical Signal Processing and Control, Volume 58Author(s): R. Suresh Kumar, P. ManimegalaiAbstractThe most important parameters of image processing are resolution of the image and processing speed. The datasets of multimedia are compressed, which are rich in quality and quantity is challenging. This paper develops a novel approach to estimate the affine parameters of fractal texture identification, in order to minimize the complexity of computation. In this proposed NLICPFTI, a pattern dictionary is maintained to hold repeated fractal patterns. Different types of data chunks such as Fractal Pattern Chunk (FPC), Intermediate Raster Chunk (IRC), Intermediate Vector Chunk (IVC), Run Length Encode Chunk (RLEC) and Flood-fill Zone Chunk (FFZC). These chunks are used to store an image into compressed format. Experimental on standard images illustrate that our approach gives more significant improvements in Peak Signal to Noise Ratio, Structural Similarity Index Mode, Feature Similarity Index Mode at high compression ratio.
Source: Biomedical Signal Processing and Control - Category: Biomedical Science Source Type: research