Design Of Visual Inspection System For Mobile Shell Surface Defects

In the era of mobile Internet,with the rapid development of smartphones,tablet and other mobile terminals,consumers are increasingly demanding the quality of these products.Product appearance is an important indicator of product quality.At present,the detection of mobile phone shell surface defects still uses manual visual inspection.This detection method has problems such as high labor costs,subjective influence of testing standards,and non-strict quality control,which seriously impedes the production efficiency of the product.In recent years,machine vision inspection technology has developed rapidly.Its features of high efficiency,stability,non-contact,and quantification have laid a theoretical foundation for the detection of mobile phone shell surface defects.In this paper,the surface defects of the mobile phone shell as the research object,combined with machine vision detection methods,research and design of mobile phone shell surface defect detection system.This article mainly carries out the following work:(1)Build a hardware platform for detection systems.The optical properties of the surface defects of mobile phone shells were studied.After many photo-taking experiments,the imaging equipment was determined.According to the spatial position parameters between imaging devices,the mechanical structure is designed.Select the motion control module and design the hardware circuit diagram.(2)Research and implementation of mobile phone shell surface defect detection algorithm.Based on the hardware platform selected and built,the image is captured,and the defect gallery of the surface of the mobile phone shell is stored.The effect of the detection is studied and analyzed through a commonly used image processing algorithm,and the relevant algorithm is designed.Just as follows,aiming at the target position of the mobile phone shell,template matching and affine transformation are used to intercept the target area and remove the region of interest,so as to reduce the computational complexity for subsequent algorithms.For scratch defects,a method based on Gaussian line detection was used to adjust Gaussian kernel function parameters and scratch defects were extracted.For pit defects,the method is mainly based on frequency domain analysis,which effectively enhances pit defects,and combines Blob analysis methods to extract pit defects.(3)System software design and implementation.Based on the mobile phone shell surface defect detection platform,software development and system debugging.The system software mainly includes: PLC-based motion control program design,camera driver program design based on Dalsa Sapera LT,implementation of defect detection algorithm based on HALCON algorithm library,human-computer interaction interface design based on MFC.On the system platform set up,the system test experiment was conducted and the defect inspection information was collected.The results showed that the detection accuracy rate of this system can reach 96%,and the low miss detection and false detection rate can reach 0.12 mm,and the surface of the mobile phone shell can be realized.Defect detection,the system has practical and practical significance.