Font Size: a A A

Design Of Workpiece Size Measurement System Based On Machine Vision

Posted on:2018-10-27Degree:MasterType:Thesis
Country:ChinaCandidate:F J LiFull Text:PDF
GTID:2382330542484225Subject:Engineering
Abstract/Summary:PDF Full Text Request
Car key as an important part of the car,the detection of its size is an indispensable part of the production process.Car key is tooth tape with internal groove,the structure is complex which need to detect a lot of points,it is hard to use traditional measurement to detect accurately.This paper combined a orizontal subject,emphasizing on the problems existed in detecting the car's key,developed a set of dimension measuring system for automobile key based on machine vision.The main contents of this paper included:Firstly,this paper discussed the background and significance of the dimension measuring system for automobile key based on machine vision,and looked through the research literature at home and abroad,and gave the overall design plan and technical route of the measurement system.Secondly,according to the design requirements of the system,completely achieved the design and transformation of the hardware system,setting up the measuring device for automobile key based on machine vision.According to the detection process of the industrial field to design the software of the system,which mainly included three parts: image acquisition,image detection and product sorting.A set of size measurement systems based on automotive key artifacts was designed.Thirdly,using the image processing technology to correct histogram,denoising,sharpening,threshold segmentation,edge detection,contour extraction,edge refinement,and other analysises and comparisons on the key which was going to be detected so that pick up the key contour image with clear edge.This paper proposed the target area of roi location picking up technology,set the area of key was only selected by roi location,and other non-roi locationed areas were set with gray value the same as background,in order to effectively reduce the amount of data processing,and divided the inner contour split into two parts,searching ans pick up the characteristic detecting point in the in profile.Finally,the key profile studied in this paper was a complex irregular shape,which involved straight lines,semicircles,curves,etc.,the points which need detecting were not located in the contours of the inflection points,some of them distributed in the midpoint of the line,using traditional hough straight line detectingmethod and least squares fitting method was difficult to accurately pick up the feature detection points.This paper used the LSD linear detection algorithm to pick up and detect the key formed with complex structure.When after using the LSD algorithm to detect the upper contour,we could get the gather of the endpoint and the line so that it designed the endpoint set sorting algorithm,the segment size sorting algorithm,the midpoint coordinate algorithm,the longest line search algorithm,in the gather of the endpoint and the line to search and pick up the location information of 8 characteristic detecting points,completes the measurement of the eight stations.Then use many samples for measurement test,the measurement accuracy and missed rate were analyzed,the design of the entire measurement system performance indicators were evaluated.
Keywords/Search Tags:Machine vision measurement, Picking up region of interest, LSD line segment detection, Image segmentation, Threshold segmentation
PDF Full Text Request
Related items