Research On Image Acquisition And Processing System Of Rigid Guide Based On ZYNQ

The vertical shaft hoisting equipment is the main form of coal production.As the guide device of the lifting vessel,the rigid guide is an important part of the vertical shaft hoisting system.The health state of the rigid guide determines the stability and safety of the lifting vessel during the lifting process.Due to shaft deformation,installation error,friction and impact damage and other reasons,the rigid guide is prone to deformation and bending and other defects,which poses a major threat to the safe operation of lifting vessels,and even causes malignant accidents such as jamming guide.The traditional post post and scheduled maintenance methods are not only high cost,low efficiency,poor real-time performance,but also susceptible to human subjective factors.In this paper,machine vision is used to monitor the health state of the rigid guide.It is difficult to realize the high-speed processing of real-time image acquisition and processing system because the current system based on DSP or single FPGA is insufficient in the aspects of capacity,computing power and embedded soft core performance.Therefore,this paper studies and designs a ZYNQ-based image acquisition and processing system for rigid guide.The image acquisition and processing system of rigid guide proposed in this paper adopts the image acquisition and processing and recognition scheme of CMOS+FPGA+ARM,which solves the shortcomings of the system capacity,computing ability and embedded soft core performance,thus realizing the health monitoring of rigid guide.Use OV5640 camera,ZYNQ-7000 extensible platform,HDMI display,LED lights of rigid guide health status monitoring system rigid guide image real-time acquisition,and the rigid guide the fault pattern recognition,image identify rigid edge guide line of two parameters,namely the plane rectangular coordinates and the distance to the origin of coordinates straight line distance to the origin of attachment and x axis Angle,through rigid guide and health condition compares two parameters at the edge of the straight line,achieve the goal of rigid guide fault identification,Finally,the health status of the rigid guide is displayed in real time on the HDMI display.Firstly,the rigid tunnel image was preprocessed by median filtering,Sobel edge detection,image binarization,image segmentation,etc.MATLAB was used for verification.The edge warning model of rigid guide was built,and the image recognition algorithm was designed to recognize the five states of the rigid guide:normal,raised,sunken,inclined and misplaced.Secondly,the hardware of image acquisition and processing system of rigid guide is designed.The hardware system with CMOS+FPGA+ARM as the core is built.The system includes the selection of components,the design of image acquisition module and the design of image preprocessing system.In terms of software,the Linux operating system was built,the straight line detection algorithm of the edge of the rigid guide was designed,the upper computer interface of the health monitoring of the rigid guide was designed,and the health state monitoring of the rigid guide was realized.Finally,the ZYNQ-based rigid guide image acquisition and processing system is tested and verified experimentally.Five failure modes of rigid guide passage are simulated in vertical hoist test bed,and the image acquisition and recognition are carried out by the image acquisition and processing system of rigid guide passage based on ZYNQ.The results show that the system can accurately identify the rigid guide passage faults and display them in real time with an accuracy rate of 99%,which has high engineering application value.There are 46 figures,12 tables and 94 references in this paper.