基于图像识别的物料分拣系统设计

北京理工大学珠海学院2020届本科毕业设计Design of material sorting system based on image recognitiondesign of image recognition and processing systemAbstractWith the continuous development of science and technology,the factory graduallystepped into intelligence and automation.Material sorting is an indispensable process inproduction.The usual manual method of sorting materials is too cumbersome and wastesmanpower and time.Therefore,we design a mechanical system for material sorting withimage recognition instead of manual operation.This article mainly describes the design ofimage recognition and processing system and the design of the circuit.Using Matlab toprocess and recognize the image,the image processing should first collect the material image,then use Matlab to gray the image and segment the image;the image recognition programuses the algorithm provided by Matlab to design,training the processed image with CNNneural network and the characteristic templates of the materials are stored well.Finally,thematerials captured by the camera are compared with the stored characteristic templates forrecognition and classification,and the data of the output material class is transmitted toArduino.This design is mainly used in the production of factories,but at present somefactories still use manual sorting method,and PLC controlled sorting system is common inthe market,from the result of recognition,it can show what material is more directly,so thispaper designs a material sorting system based on image recognition.Keywords:Image recognition;Image processing;Matlab;CNN neural network;Arduino目录1绪论1.1本设计的国内外发展历程与现状1.2本设计的研究意义与目的…2