关键词: 粒子群优化, 雁群, 线性递减惯性权重, 直方图, 熵

Abstract: Image segmentation is a key part in image processing field.At present，there are several image segmentation methods，among which the thresholding method has predominant advantages.But the key of the thresholding method is to find the optimum entropy threshold of an image effectively.To solve the problem，the location updating equation of the Geese-LDW-PSO algorithm has improved that the present position of all particles is replaced by the global best value of the population.And the improved algorithm is applied into entropy thresholding image segmentation method.The simulation results show that this algorithm can obtain the optimum threshold value of an image rapidly and stably and has good performance in the segmentation of a vehicle brand image.

Key words: particle swarm optimization, wild geese, linear descend inertia weight（LDW）, Histogram, entropy