Image segmentation based on two-dimensional histogram and the Geese particle swarm optimization algorithm

Abstract

Image segmentation is a key part in image processing fields. The image segmentation method based on maximum entropy thresholding and two-dimensional histogram has many advantages, but it requires a large amount of computing time. To solve this problem, the Geese-LDW-PSO algorithm was introduced in this paper. Here, the Geese-LDW-PSO which was inspired by the wild geese group was the particle swarm optimization attached with linear descend inertia weight. First, the Geese-LDW-PSO was used to seek the optimal threshold value of a picture adaptively in the two-dimensional gray space. Then, the picture was segmented with the optimal threshold value which had been gotten. The simulation results showed that the Geese-LDW-PSO algorithm performed better in the segmentation of a vehicle brand image.