Improved Particle Swarm Optimization Crack Identification for Building Wall

Aim at the problem of low detection accuracy caused by irregular crack shape and complex background in wall crack detection scene, an improved particle swarm optimization image segmentation algorithm is proposed to identify the crack images. In the first place, the threshold segmentation problem is solved by using improved dynamic inertia weight and mean position update method. At the same time, Otsu algorithm is combined with filtering and morphological processing in our new algorithm. Experimental results show the algorithm proposed in this paper has well effect on the detection of wall cracks that compared with genetic algorithm and the other two particle swarm optimization algorithms. Not only that, but the experiment also shows the proposed approach has well segmentation results for other images.