Design And Implementation Of Full Appearance Visual Inspection System For Rotary Parts

During the production process,parts will inevitably have various defects,such as dimensional deviation,scratches,scratches,and surface contamination.These defects can affect the product to varying degrees,such as some precision parts in automotive engines.Defect detection of parts is an important means to ensure the safe use of parts.Traditional manual detection methods not only have low detection efficiency,but also have large randomness.It is difficult to ensure consistent quality control.Machine vision inspection technology is widely used in the inspection field due to its non-contact,high efficiency and high precision.In order to meet the needs of full-surface defect detection of rotating parts,this topic studies the inspection system based on machine vision to realize the automatic detection of full-surface defects of parts.The main research contents are as follows:1.Hardware system design.According to the shape,material,defect type and inspection requirements of the rotating body parts,design a hardware system that can highlight the performance of the defect,including the overall structure design of the system,the choice of camera,optical lens,light source,and lighting method.Through the lighting experiments,the spatial combination of the parts,light sources,and cameras corresponding to various defects is determined to ensure that the inspection system can obtain a clear image of the defects.2.Research on image processing algorithms.This paper proposes a method based on machine vision for full-surface defect detection of rotating parts.This method uses wavelet denoising for image preprocessing.Based on Otsu algorithm,the problem of double threshold selection in Canny operator is improved,and the image is well realized.Defect segmentation.Hu moments are used to extract defect features and SVM classifiers are trained by sample features to identify and classify the appearance defects of parts.3.Software system design.This system is based on Lab VIEW software and NI Vision module,and uses the built-in MATLAB script control in the software to meet the requirements of full-surface defect detection of rotating parts.The system uses a modular programming method,which is individually programmed for each step of image processing.These include image preprocessing,image segmentation,feature extraction,and pattern recognition.The processing of the defect image of the rotating parts is realized.In order to meet the requirements of full-surface defect detection of rotating parts,this subject designed a hardware system and a software system to implement full-vision defect detection of rotating parts based on machine vision.Taking the actual part defect image as an example,the experimental results and algorithms are analyzed,and it is verified that the system can effectively solve the problem of full appearance defect detection of rotating parts.