Research On Image Processing Algorithms For Steel Core Conveyor Belt Defect Detection System

Steel cord conveyor belt is a major coal transport equipment used in coal mining enterprises have become increasingly popular, but often because of its breaking of coal production caused by a defect rope interrupted. Therefore, in order to ensure safety in production, to prevent fatal accidents, it is necessary to detect defects in the case of steel cord conveyor belt steel cord.In this paper, the main work done as follows:In this paper, the linear X-ray image sequence capture arrays run steel cord belts, according to its texture features, a new steel cord image texture coding defect detection algorithms and on VC++platform to achieve the steel cord conveyor belt line defect detection and storage targets. Detection principle of the algorithm is to image gray value of the steel cord into the computer pretreatment, according to X-ray detection principle, linear array detector to collect different defects and non-defect image gray value in the form of visual images of different brightness texture nondeficit law alternately bright and dark images, while the defective part has upset this regularity, this paper is the use of texture features to detect belt defects. Texture coding is a binary linear pattern texture coding operator, first image coding lateral and vertical coding, and then calculate the transverse and longitudinal differential difference, according to the results marked defect line, details width and height defects and other defects, defect record the emergence of regional, real-time image storage defect detected. The detection system by VC++development environment, the preparation of the defect detection module algorithm code. Then create an image database, using integrated simulation laboratory, the image of a large number of steel cord experimental results show that the algorithm can be applied to X-ray the steel cord conveyor belt detection system, steel cord conveyor belt line automatic detection of defects The algorithm detects the average up to 88.44% accuracy.In addition, in order to facilitate the user to view the joint and defect detection image, culminating in the detection of security in the form of an electronic version of the report PDF document, and with Harris corner detection principle defect of the label. Harris corner detection is the image gray corner detection algorithm most commonly used, we use typical Gaussian function calculates the x direction and the y direction of the image derivative, calculated for each point of the input image partial autocorrelation matrix, based on eigenvalue, extracting the image corners, find defects in steel cord conveyor belts, and the use of bright rectangle be labeled. This algorithm experimental data collected from the field, true and reliable. Needs analysis phase, personally to the scene to collect customer needs, debugging stage, in addition to the development of the use of this simulation laboratory, but also often to the site to debug, communicate directly with the company.Texture coding of steel cord image defect detection algorithms to identify the defective area is not only to meet the online real-time detection, and the algorithm is simple, less time consuming, the study is of important theoretical significance and value...