改进SSA的搪瓷盘表面瑕疵检测研究OA

In order to realize the automatic detection of surface defects of enamel plates,a method of enamel plate defect detection based on small samples is proposed.A platform was built to acquire the enamel plate images.The images were processed and the RF algorithm was used to filter out the high-quality features as feature selection.The SSA(Salp Swarm Algorithm)was introduced to improve the algorithm and optimize the support vector machine and build the TSSA-SVM detection model.The experiments show that the TSSA-SVM detection model has better effectiveness and stability compared with other methods.The proposed method for detecting enamel disc defects provides a new solution to the surface defect detection problem of enamel discs.