Defect Detection Of Graphite Heat Sink For Mobile Phone Based On Machine Vision

As mobile phones are widely used in society,the demand for graphite heat sink of mobile phones is also growing rapidly.Although the production of mobile phone graphite heat sink is large,it is inevitable that unqualified products will appear.Most enterprises generally use manual detection method.In the process of detection,there are problems of low detection accuracy and high missing rate.Therefore,a fast and efficient method is urgently needed to improve the accuracy of detection of mobile phone graphite heat sink and reduce the rate of missing inspection,so as to reduce the cost of products.Therefore,this thesis studies a defect detection system of mobile phone graphite radiator using machine vision.Through the research system,we can realize the online automatic detection of mobile phone graphite radiator,and complete the defect classification in the image of mobile phone graphite radiator.The research contents of this paper are as follows:(1)Design of defect detection system for mobile phone graphite radiator.According to the characteristics of mobile phone graphite heat sink and the requirements of production inspection,the hardware design of the system is completed,including the selection of camera,lens,light source and other components for image acquisition,and the experimental platform is built;the user interface is designed by using Visual Studio software,and the defect detection of mobile phone graphite radiator is completed by using Halcon visual software.(2)Research on image preprocessing and image matching of mobile phone graphite radiator.Firstly,several filtering methods are compared and analyzed,and a homomorphic filtering method based on adaptive local noise reduction is proposed.This method can get higher contrast images than traditional filtering methods.(3)In this thesis,the image template matching method is studied to make the image positioning.The gray template and shape template are mainly used to match the image of the heat sink to be tested and the template image.Due to the instability in the process of loading and transmission,the image position will be tilted or shifted.Therefore,before the image matching,the image geometric transformation is carried out,and the image position of the mobile phone graphite radiator is corrected by translation and rotation Compared with the image matching experiment,the shape template matching is better for image location.(4)Defect extraction of graphite radiator for mobile phone.The defect regions are displayed by differential operation method,the noisy spots are removed by morphological method,and the defect features are extracted by Blob Analysis(Blob Analysis).(5)Software and hardware joint debugging of mobile phone graphite radiator detection system.After the implementation of the software algorithm of mobile phone graphite heat sink,the software system combined with the hardware part of the system to test the system functions to verify the stability and reliability of the system designed in this project.The experimental results show that the mobile phone graphite radiator detection system designed in this thesis can meet the requirements of enterprise quality control.