Crack image detection based on LS-SVM optimized by PSO

A new crack image detection method based on Last Squares Support Vector Machines (LS-SVM) optimized by Particle Swarm Optimization (PSO) is proposed. Firstly the image intensity of neighborhood region of pixel is well estimated by LS-SVM, the gradient operator and zero crossings operator are obtained by LS-SVM based on Gaussian kernel. Then the method of edge detection based on the combination results of gradient and zero crossings is used to crack image detection. The parameters of LS-SVM are optimized by Chaos Particle Swarm Optimization Algorithm (CPSO) in order to achieve best edge detection performance. Finally, crack detecting experiments of glass bottles validate the effectiveness of the proposed method.