| In order to improve productivity and reduce the error caused by manual operation,the development of automatic coil delivery system has become a necessary requirement for the process of cold rolling line.The traditional automatic coil delivery system includes diameter measuring method,encoder method,ultrasonic positioning method and so on.The controllability of these systems is not strong,and too many sensors lead to high failure rate and difficult maintenance.In order to solve this problem,this thesis designe an automatic coil delivery system of uncoiler based on image processing,and propose a coil location algorithm apply to this system,to realize the automatic control of the coil delivery process.The system includes sensor module,communication module,processor module,controller module and vehicle motion control module.The image algorithm in this thesis transforms the problem of coil positioning into ellipse detection,which is divided into two parts:target search and ellipse high precision detection.In the target search algorithm firstly,search the initial ellipse target in the static region of interest,and then describ the image preprocessing algorithm in detail.Compare the Sobel algorithm,Canny algorithm and Canny algorithm edge detection method with adaptive parameters respectively,and analysis of the ellipse detection algorithms SHT,RHT and three-point method in detail in this system.In the high-precision detection algorithm of ellipse,putting forward the dynamic finite domain method,the dynamic ellipse search method based on the long and short axes,the contour separation method and the quadrant method for filtering,then fit the samples by the least square method.Put forward the dynamic ROI method to achieve target locking and tracking,and obtain the position coordinates and shape parameters and obtain the position coordinates and shape parameters of the ellipse,which is equivalent to the positioning of the steel coil.Through the comprehensive analysis of the data,fitting the central motion equation of the coil,the long axis,the short axis and the inclination angle of the ellipse to design the reliability evaluation module,and the reliability of the test results is measured by 0%(Not credible)to 99%(Highly reliable),then the system decides whether to use the automatic control instruction according form the reliability,ensure the safe operation of the system.The industrial field verification shows that the system has high accuracy,good real-time performance,low failure rate and stability under the condition of sufficient illumination.The detection system has stronger controllability and is easy to maintain in the later period,which improves the efficiency of industrial production. |
