Research On Design And Key Technology Of High-Speed Vision Inspection System For Feeding Belt

The feedingbelt is a common packaging coil in the microelectronics industry.More and more carrier manufacturers use various automatic detection devices in the manufacturing process to ensure the reliability of the feedingbelt quality.The traditional method of detecting is mainly artificial detection,but the existence of detection speed is slow,the detection accuracy is low,can not guarantee the reliability of manufacturing,can not meet the requirements of modern manufacturing.In this paper,some common defects of the carrier tape are used as the detection object,and the key technology of rapid detection and identification of the defective detection system of the carrier tape with on-line detection is studied systematically.In order to replace the artificial Detection,the main research content is as follows:(1)Detection system design.According to the requirement of on-line detection,the structure,working flow and detection scheme of the detection system are determined.The light source,camera and lens and image acquisition card of the system hardware are determined according to the requirements of the defect detection accuracy.Alsotheplatform HALCON.(2)A Fast Location Algorithm for Image Feature Area.Based on the high quality image of the carrier,the image is preprocessed.This process includes the gray scale,median filter and threshold segmentation of the image,and then introduces the main steps and basic principles of image splicing.Based on Harris algorithm This paper introduces the calculation method of two kinds of image feature regions: based on the calculation and analysis of the cumulative distribution of the image gray value and the fast algorithm of calculating the one-dimensional moment of the image,Based on the comparison between Harris algorithm and the algorithm adopted in this paper,the results show that the proposed algorithm has improved the splicing speed and matching speed.(3)Rapid Detection and Identification of Defects in Feed.Which is divided into template matching method based on shape feature and detection method based on gray scale analysis.The template matching method based on shape feature is based on HALCON to match the target image,and then the image is different.The basic principle of the three template matching techniques based on HALCON,the basic principle of the image difference method and the main process of the algorithm are described.Analytical detection methods,including image grayscale processing,image histogram analysis and Blob analysis.