Research On Machine Vision Tube Stain Detection System

With the rapid development of the textile industry,product quality control has become the focus of research in the industry.Among them,the yarn tube plays an important role in the industrialization of textile industry equipment,and the quality problem of the yarn tube has gradually attracted the attention of the textile industry.In view of the current traditional methods of manual detection and machine inspection,which cannot meet the requirements of the rapid development of modern textile industry,this thesis studies the detection technology of tube stains from the improvement of the accuracy and real-time of the detection of tube stains.This technology has important implications for the development of the textile industry.This thesis combines machine vision,image processing and other technologies with the textile industry.Based on the existing platform of the school laboratory,this thesis develops and designs a set of yarn tube stain detection system using image processing technology to realize the stain detection of the tube during the handling process.The main work and contributions are listed as follows:(1)Based on the characteristics of the textile tube production process and the needs of the project,the design goals and difficulties of the research system are studied.The overall design of the tube stain detection system is mainly composed of the overall design scheme of the system and the system design selection of hardware and software.(2)Design of yarn tube pretreatment algorithm.It mainly includes image graying,image denoising,image binarization,standard template making and image registration processing.Through experimental analysis,this thesis chooses weighted average method for grayscale,fast median filtering for noise reduction,maximum inter-class variance method for binarization,statistical average method for standard template,and SURF feature point matching algorithm.Registration of the tube image in spatial position.(3)Design of yarn tube stain detection algorithm.Based on the pretreatment of the tube image,this thesis uses image differential,layered detection and template matching to analyze the three commonly used stain detection algorithms,and proposes to detect the stain based on the gray histogram template matching method.Analysis of the gray value and pixel points in the gray histogram can objectively detect the tube stain.(4)According to the system design requirements,use Visual Studio and Halcon software to design an easy-to-operate PC interface.In the laboratory environment,the yarn tube stain detection algorithm was realized,and the development of the detection system was completed.In this thesis,the yarn tube stain detection algorithm and system are tested.The test results show that the real-time and accuracy can meet the design requirements of this thesis.The accuracy of yarn tube stain detection is over 95%,and the feasibility is strong.The development of detection technology has certain application value.In the text,there are 42 pictures,13 tables,and 58 references.