Research Of On-line Glass Defect Inspection System Based On Digital Image Processing

With the development of society and the improvement of science and technology,there is a greater demand for a better quality of glass products.The traditional detection and recognition methods can not meet the requirement for real-time online detection and the need for high quality any more.The system of on-line glass defect inspection is an important way which can improve the quality of glass products,reduce scrap rate and improve the efficiency.So the on-line glass defect inspection system based on digital image processing is designed in this paper.In order to extract the feature vector of glass defects and recognise them,the system combines the improved moment invariants algorithm and radial basis kernel function of support vector machine algorithm.Firstly,the overall scheme of on-line glass defect inspection system based on digital image processing was designed,and the composition and function of each module is intoduced in detail.Secondly,the characteristics of glass defect of traditional extraction algorithm was discussed.According to the characteristics of glass defect,the moment invariants algorithm is put forward.Based on the algorithm experiment of glass defect extraction and defect matching is done.The results show that the algorithm can accurately match to the target image and can also display the coordinates of the location of the target image accurately.The experiment also display that matching accuracy can reach 98%and the efficiency is increased by 34%compared with the traditional algorithms.It can meet the requirement of on-line glass defect detection.Finally,the support vector machine algorithm of glass defect classification based on the RBF kernel function was put forward.The efficiency and accuracy rates of the support vector machine classification algorithm are higher than those of the traditional algorithms.Through comparison experiments,the identification rate of RBF kernel function support vector machine algorithm was significantly higher than that of other kernel function algorithm.The overall recognition rate of RBF kernel function of support vector machine algorithm reached 95%,and it can meet the application of on-line glass defect inspection.