Multi-threshold Image Segmentation Through an Improved Quantum-Behaved Particle Swarm Optimization Algorithm

摘要

Multi-threshold segmentation is a basic and widely used technique in image segmentation. The key step of accomplishing this task is to find the optimal multi-threshold value, which in essence can be reduced to multi-objective optimization problem. The quantum particle-behaved swarm algorithm (QPSO) is an effective method to resolve the problem of this class. However in practice, we found the original QPSO has imperfections, such as the excessive dropping of the diversity of the population and trapping in local optimum. In order to improve the ability of searching the global optimum and accelerate the speed of convergence, we proposed an improved quantum-behaved particle swarm algorithm (IQPSO). The experiments showed that IQPSO was superior to PSO and QPSO on the searching of multi-threshold value in image segmentation under the premise of ensuring the accuracy of solutions.