The Design Of Surface Quality Of Plastic Setscrew Based On Machine Vision Online Detection System

plastic setscrew (also called plastic inflation screw connections) can beshape changed and have defects on the surface during plastic injection molding orassembling process influenced by mechanical vibration and complex productionenvironment. This dissertation applies machine vision to the surface detection ofplastic connections. According to the assembling process planning image detectionprocess to design image processing algorithm based on the defect type and feature;Designing a mechanical sorting device combined with assembling process; at last,Designing a detection control software to achieve100%detecting, then controllingthe sorting device to remove defective products based on surface quality analysisresults.To guarantee the requirement of inspecting accuracy and speed on the design ofmachine vision detecting system.Meet the requirement of detecting speed and theoptimize of algorithm increases both of detecting accuracy and speed. Thisdissertation can be divided into four parts as followed:(1)Technical route and detecting method: Researching on the machine visioninspecting system and image processing algorithm at home and abroad, which isbased on the technical route of surface quality detecting about background,significance and application. Choosing machine vision detecting method instead ofoff-line spot check, using external triggering(triggered by impulse signal)USBindustrial camera to achieve100%automatic on-line detecting, the camera is alsovery cheap.(2)Integral inspecting scheme design: According to the assembling process ofplastic connections and mechanical structure of production line, designing theintegral scheme, including image processing software, algorithm, image capturedevice, mechanical sorting device and control module design, etc.(3)Image processing algorithm design: Due to the available filtering algorithmshave defects of fuzzification on the image edge feature during filtering pattern noise, this dissertation gives a kind of filtering algorithm called rotating mask, which canfiltering pattern noise well and get the sharpening result on the image edge feature.Adopting dual threshold Canny edge detecting method based on image edge feature,and using image segmentation after filtering. During the image matching, designingthe multi-template matching process, this has anti-rotary circle template matching,and combined with Zernike matrix matching.(4)Inspecting protection and sorting device design: To ensure the long-termstability running of the whole system and convenient debugging in the later period,designing inspecting protection device, capturing stable image, reducing theinfluence of high-frequency vibration on the processing of assembling as little aspossible. The sorting device using the stable air source, which is available on theproduction line, including swinging cylinder and qualified, detective slide groove. Toachieve automatic sorting by the command of controllerThis design of plastic connection surface quality detecting system is based onmachine vision, the image processing algorithm can meet the requirement ofdetecting goal, which can realize automatic image capture and detecting the targetsurface quality under the production speed of120pieces/min, the wrong detectingrate is below2%.