FicBit: An improved Fractal Image Compression using Bio-Inspired Techniques

With the surge of modern communication technologies, image compression has attracted significant interest in the research community. Fractal image compression (FIC) is regarded as one of the prominent compression methods, due to the high compression ratio and short decompressing time. However, one major drawback of this approach is its high computational cost, which results in a long compression time. In this work, we propose FICBIT, an improved FIC technique leveraging numerical properties of fractional encoding and binary conversion system. We augment the fractional-to-quotient number-based image compression technique by performing channel splitting and optimal sub-imaging. We further optimize the compression by leveraging efficient meta-heuristic algorithms, specifically genetic algorithm (GA) and particle swarm optimization (PSO). We introduce a novel generalized fitness function for both the algorithms and experiment with state-of-the-art image processing data-set for validating the proposed strategy.