Research On Quality Defect Detection System Of Cutting Blade Based On Image Processing Technology

Cutting blades are widely used in cutting ordinary steel,stainless steel metal and non-metallic materials,due to its high tensile,impact and bending strength.In the process of production and preparation,the quality defects of the cutting blades are caused by various reasons.The defective inferior products should be removed from the production line through inspection.There is demand for manual intervention to complete the detection presently,but the low efficiency and high false detection rate makes it difficult to meet the requirements of modern production automation manufacture.This paper present the flaw detection method based on image processing that can realize disfigurement real-time,rapid and accurate detection,which is helpful for promoting the technology of automation detection.Firstly,the image processing platform,camera,lens and light source are selected according to the size and shape of the cutting blade.The ring light source with a reasonable angle is used for targeted lighting to obtain high-quality cutting piece images.The target image is extracted from the background image,and the center coordinates of the circle are obtained by fitting the edge contour to realize the regional segmentation of the cutting blade.Secondly,the registration of standard map and pending map is matched based on SURF algorithm combined with affine transformation.The efficiency of matching is improved by selecting feature regions to match and the accuracy by removing mismatching through Euclidean distance constraint and RANSAC algorithm.Finally,based on the difference image processing method and Blob analysis method,the defect result is extracted from different regions,and the minimum outer rectangle of the defect is calculated,which is drawn in the graph.Based on the effective image acquisition hardware platform,the detection algorithm was written on the Matlab platform to verify the effectiveness.Then,writing C++ program code on the Visual Studio platform calls Open CV to write the defect detection program for cutting slices.Both of the algorithm efficiency and time efficiency are optimized.Real-time online detection system is able to realized disfigurement by calling image acquisition camera and its processing function SDK of hardware platform.