Research On Cable Strand Quality Detection Based On Machine Vision

Wire and cable is an indispensable accessory product for the national economic sector.From ultra-high voltage transmission lines to various micro-motors,people can't live without wires and cables in every aspect of their production and life.Wide variety of wire and cable,large quantity and wide range.Many varieties are also included in the national product safety certification product range.In general,the quality level of domestic wire and cable is obviously lower than that of foreign countries.Thousands of domestic manufacturers,users involved in all walks of life in various systems,a large number of manufacturers produce inferior wire and cable.Because the problems of wires and cables cause fires or other accidents that occur from time to time,the detection of cable strands is very important.In this paper,machine vision technology is used to detect and research the quality of cable stranded wire,and a set of feasible quality inspection scheme is formulated.This program is based on Windows system,C++,Microsoft Visual Studio 2010 development tools,MATLAB as an auxiliary research tool,using OpenCV computer vision processing library for detection algorithm and interface design,and ultimately solve the cable strand detection problems.In this paper,according to the quality problems that often occur in the production line of cable twisted wires and the international standards for qualified cable lines,the quality inspection algorithm for cable lines is studied.At the same time,various testing algorithms are tested to determine the appropriate algorithm.At the same time,a design proposal based on the quality inspection of machine vision cable strands is also proposed.First of all,in order to collect high-quality pictures,we need to configure the light source and build a good environment for image collection.Then multiple cameras are called to perform multi-directional image acquisition of the cable strands and transmit them to the PC via the data cable.Then the image is filtered to eliminate noise,erosion,swelling,Canny edge detection,contour extraction and other processing.Then use the minimum distance method,least squares method,Hough transform and other linear detection algorithms to extract the surface of the cable strands.Finally,using the oil pollution detection algorithm,the detection of lanterns,the detection of the diameter,the detection of the broken line,and the detection of broken lines,the detected images are detected,and the purpose of on-line detection of the cable strands is finally achieved.In addition to the three-dimensional reconstruction of the surface of the cable strands,a certain amount of results have been achieved.This article mainly completes the online detection of common jumpers,loose strands,oil stains,broken strands,and diameters in the cable stranding line.By simulating the industrial environment,the cable strands were tested.Experiments show that this solution can basically solve the problem of on-line quality inspection of cable lines.