Research On Detection Method Of Ring Spining Broken Yarn Cruise Driven By Machine Vision

Ring spinning accounts for more than 80% of the total production in the field of spinning with its advantages of wide variety of production,strong adaptability of raw materials and good spinning quality.In the ring spinning process,there is a phenomenon of broken yarn,and the failure to timely obtain the broken yarn information to remind the staff to connect the yarn will lead to idle spindle position,affecting the production efficiency and product quality.So it is of great engineering value to study the automatic detection method of ring spun yarn break and feed back the information of broken yarn in the spinning process in real time,so as to remind the staff of timely joint and improve the working efficiency.The traditional single spindle testing is difficult to be applied to the existing spinning frame due to its high cost and difficult transformation.Therefore,this paper studies the roving detection method of ring spinning broken yarn.Aiming at the weak yarn feature in spinning image under high noise,the smoothing and enhancement methods of spinning image are studied.Aiming at the problem that it is difficult to identify broken yarn under small target,the characteristics of medium frequency domain and space domain of spinning image are analyzed to accurately identify broken yarn.The main research work is as follows:To improve the contrast between yarn and background in spinning image,this paper studies the spinning image enhancement method under high noise.Aiming at the difficulty of weak yarn characteristics under high noise,a spinning image smoothing method based on L?0norm was proposed in image processing to suppress background noise.In feature processing,a gradient feature extraction method based on neighborhood enhancement was designed to enhance the neighborhood gradient feature of spinning image.The results show that the weak feature two-stage enhancement algorithm can enhance the spinning image and enlarge the difference between the yarn and the background.In order to realize the detection of broken ring spinning based on vision,an image-based identification method of broken ring spinning is studied.Otsu adaptive threshold segmentation method is used to solve the problem that yarn is susceptible to light.Aiming at the difficulty of small yarn target in spinning image,a threshold search method inspired by valley bottom was designed to extract yarn features.On this basis,the identification method of broken yarn based on Euclidean distance is designed by taking advantage of the same space between each spindle in ring spinning.The results show that the small target image broken yarn recognition method based on frequency domain and space domain can effectively identify the broken yarn in the image and realize the location of the broken yarn.Since the above theoretical research,the ring spinning broken detection microservice system was built to realize the high integration of spinning image enhancement and broken detection,which verified the effectiveness of ring spinning broken detection under visual features.The research results of this paper provide theoretical methods and technical tools for ring spinning broken yarn detection,which is of great engineering value for ring spinning to realize low-cost automatic detection.