Discrete Quantum-Behaved Particle Swarm Optimization for 2-D maximum entropic multilevel thresholding image segmentation

This paper presents an improved discrete quantum particle swarm optimization (IDQPSO) for 2-D maximum entropic multi-threshold image segmentation algorithm. Firstly, particle swarm binary-encoded method based on 2-D threshold is proposed. Additionally, new particle evolution strategy is proposed to avoid converging on local optimum and accelerate searching progress. Additionally, experiments are conducted by comparing IDQPSO with other state-of-the-art methods such as QGA, NBPSO and BQPSO. The results show that IDPQSO outperforms other algorithms at precision, efficiency and stability.