Research On Yarn Quality Detection Based On Image

Yarn is one of the important necessities in people's daily life.The quality of yarn is related to production efficiency and value of yarn.The traditional yarn quality inspection methods have many shortcomings,and now they do not conform with the demand of intelligent production.With the development of computer and the technology of image processing,reliable,efficient and high precision digital image technology has been widely used in industrial production.How to use image processing technology to overcome the shortcomings of traditional detection methods and improve detection efficiency has become the focus of research.This paper designs an intelligent detection system based on image.In order to improve the precision of measurement,the camera is calibrated firstly,and the distortion of camera lens is solved.In order to solve the interference factors such as noise and jitter,a series of preprocessing work has been done,including grayscale transformation,tilt correction,histogram equalization and so on.Then,this paper analyzes the characteristics of different filtering algorithms and selects the most appropriate bilateral filtering to remove noise interference.After solving the problem of image noise,we need to focus on yarn samples.Therefore,this paper uses threshold segmentation method to separate the target from the background,highlighting the research focus.After the threshold segmentation,there are still isolated small holes in the image which affect the subsequent feature extraction.In this paper,combined with the characteristics of corrosion and expansion,mathematical morphology closed operation is used.In addition,the hairiness shape of yarn needs to detect the edge.Five detection operators are considered in this paper.It is found that Canny edge detection operator is not easy to be disturbed by noise,and it is easier to detect the real edge.The research focus of this paper is to detect the hairiness shape of yarn,as well as the information of coarse pitch,detail,nonuniformity and so on.Therefore,the image can be refined to facilitate the subsequent processing.In order to accurately detect the yarn quality parameters,this paper designs a pixel length matching method,which solves the problem that the traditional yarn length detection algorithm can only calculate the hairiness projection length but not the actual length,and also avoids damaging the yarn hairiness structure.This paper focuses on the study of the hairiness length of the yarn,and carries out grading statistics.It also studies the information of the yarn's gross pitch,details,nep rate and yarn line unevenness.In addition,in order to guide the actual production,the working interface of yarn quality detection system is developed.The interface can collect and process yarn image in real time,and detect yarn quality information according to the algorithm program.Experimental results show that the image method can measure yarn quality parameters efficiently and accurately,which has higher practical value.