Visual Inspection Of Quality Defects Of Small DC Motor For 3C

At present,the machine vision has a great breakthrough in the detection of standard workpiece.In the non-standard field,the universality of vision system is poor due to the lack of standard parameters.In this dissertation,the non-standard workpiece small DC motor used in the air-conditioning vane of Volkswagen is taken as the object,and the visual inspection method is studi ed for the quality defects that may occur in the automatic production line.The research work,main achievements and innovations of this dissertation are as follows.:1?This dissertation summarizes the relevant image processing algorithms at home and abroad,and analyses the relevant visual detection algorithm.For the small DC motor automatic production line with many testing stations and different field of view,the overall design of the testing system is completed.2.In image processing.Firstly,piecewise linear gray transformation based on gray histogram is studied to enhance the contrast between PIN needle image and solder image.Then,for the problem of multiple noises in workshop,the denoising algorithm based on spatial domain and wavelet transform is studied to remove multiple mixed noises.In order to extract the target area of each part,an improved maximum class variance method suitable for gear brush is studied,and the pin part of the brush is segmented.A transition area extraction method ba sed on local entropy and local variance is studied to segment the solder joint area in PCB substrate.The sub-pixel edge is extracted by Canny algorithm and sub-pixel edge fitting method,and the detection accuracy is accurate to sub-pixel level.3.In defect detection.The minimum width of irregular edge is measured by detecting the direction of edge gradient and Hough line detection.A brush deformation detection method based on parallelism tolerance is used to detect the pin offset of brush.Aiming at the problem of PIN pin position detection,the position deviation of each connected area of PIN cross section is calculated,and the difference is made with the standard value,and the result is judged by comparing with the threshold value.For the classification of solder defects such as leak soldering and short circuit,the features of binary images are extracted,and the classification decision tree is constructed to detect and recognize them.Through the four parts of the detection method design,the quality of the entire motor testing is completed.The acquisition system,image arithmetic and defect detection method of four detection parts are integrated,and a visual inspection system of small motor is developed.The results show that the detection method and the detection system developed in this paper can achieve the purpose of product quality detection and successfully put into production and operation.