Algorithm Research On Workpiece Surface Scratch And Stripe Defect Inspection

With the progress of the society and the development of science and technology, the standards of product quality are more and more strict.The surface of some products may exist defects caused by mass production of lined factories, nowadays, the defects are mainly dectected by visual identification of people.This paper takes iphone4s surface image as target.This paper is mainly research on the scratch defects and vertical stripe defects based on defect characteristics analysis and judgement of inspection.The vertical stripes defects includes vertical stripe,burr and black line defects.Through the analysis of the gray feature and projection characteristics, this paper acquires significant scratch and burr defect’s gray feature, and prominent vertical stripe and black line defect’s projection feature.Surface image exists many noise due to the slight uneven of the workpiece itself.Traditional image denoising algorithm based on smooth noise causes fuzzy edge problem at the same time.This paper has improved the TV denoising model,thus can better keep the edge information from the smooth process, and can achieve the goal of image denoising. An improved line detection template of scratch defect inspection algorithm is introduced to solve the problem of small target and unknown direction of scratch.By calculating the four directions of0degrees,45degrees,135degrees,135degrees of sequential filtering and line detection of comprehensive evaluation method, the target information is enhanced, thus can satisfy scratch inspection problem of any direction. Finally, by extract the image histogram scale binarization, and edge connection method, the weak edge intermittent problem can be prevented, thus can meet the needs of the scratch detecting. A surface defect inspectting method based on projection analysis and algorithm is proposed to deal with the problem that the existing vertical stripe inspecting methods are insufficient for high-frequency noise. A judgment method of vertical stripe defect based peak valley values is designed in order to realize automatic inspection.The paper also has designed algorithm of edge line detection based on block projection and interval defect determination algorithm based integral value by analyzing the characteristics of defects.A large number of experiments show that the algorithm can effectively identify scratchs and vertical stripes defect of workpiece surface.And the inspecting accuracy rate is nearly95%.