Research On Automobile Apparent Defect Detection System Based On Computer Vision

Automobile apparent quality inspection is one of the important processes in the process of automobile production and transportation.It has a direct impact on car sales.However,manual visual inspection is currently used for quality inspection.This method is not only inefficient,but also not conducive to saving the detection data.At present,computer vision technology is an emerging detection technology,which provides technical support and direction for automobile apparent defect detection.Therefore,this thesis intends to build an intelligent system for automobile apparent defect detection which is based on computer vision technology.The work of this thesis is as follows:(1)Based on the analysis of the system requirements of automobile apparent defect detection,this thesis designs the hardware platform and software system.We design a system acquisition scheme to realize intelligent acquisition of automotive apparent images.on this basis,we choose the hardware equipment to build the image acquisition platform.Besides,we design software systems by combining hardware and software system requirements.(2)we design the module of automobile apparent image processing.In order to select the detection algorithm of this thesis,we compare a variety of deep learningbased object detection algorithms.By collecting vehicle appearance image,we construct the data set needed for algorithm training.For the problem of image noise,we study the image denoising algorithm.In addition,we propose a robust image processing algorithm based on matrix regression to zoom in on the collected lowresolution image.Due to the limited sample collection conditions,we studied the data enhancement technology to realize the amplification of automobile image.And we use the intelligent detection algorithm to train the defect detection model.(3)This thesis implements the establishment of hardware platform and the interaction of system software functions of the automobile apparent defect detection system.We analyze and test the detection model in this thesis,and the result shows that the detection accuracy of the algorithm model in this thesis reaches 91.07%.Besides,we integrate hardware acquisition platform and image processing algorithm to develop system detection interface.We implement the functions of flaw detection,data storage and historical query to ensure the efficient and stable operation of the intelligent detection system.In short,compared with traditional manual detection,the automobile apparent defect detection system developed in this paper is not only conducive to improving the defect detection efficiency and accuracy,but also has data storage and query functions,providing data support for subsequent liability source.