Development Of Copper Strip Defect Online Detection System Based On ARM And DSP

In the production and processing of workpieces or raw materials,the quality and production efficiency increasingly become an important part of intelligent manufacturing.However,the traditional manual sampling inspection is not only inefficient,but also easy to be affected by subjective factors.To solve this problem,this paper designs an online copper strip defect detection system based on ARM and DSP to complete the detection of surface quality of copper strip on the production line.Compared with based on PC online detection,this system has higher cost performance,faster detection speed and more convenient assembly.The main research contents of this paper are as follows:(1)Design an embedded online copper strip surface defect detection system.According to the defect characteristics of the copper strip,a dark area lighting mode is designed to complete the optimal imaging of the surface defects of the copper strip.OMAPL138 which is based on the combination of ARM and DSP is used as the core processor to complete real-time defect detection.(2)Research the defect detection algorithm based on MM-Canny.By improving the Canny operator and combining it with the morphological method,the image defects can be quickly and accurately identified and marked.A defect classification model based on support vector machine(SVM)is established.And for improving software efficiency,optimize and review codes on ARM and DSP.(3)Design and implement the system software,including acquisition module,communication module,display module and image analysis module.The acquisition module is responsible for configuring the camera driver and frame mode of image processing.In the communication module,the message communication mechanism based on MessageQ and the data interaction mode based on shared memory are studied to realize the inter-core communication of OMAPL138.The image analysis module is responsible for analyzing and processing the copper strip images.Among them,DSP is used to realize defect recognition of detected images,and off-line support vector machine is used to complete defect classification on ARM.(4)Test the surface defect detection system of copper strip.The experimental results show that the accuracy of system defect detection reaches 99%,the accuracy of classification is maintained at about 95%,which is 5% higher than manual classification and the system can meet the application requirements of industrial detection.