Design Of Track Displacement And Temperature Superimposed Monitoring System With Video

With the development of railway and the wide application of seamless line,the service status of infrastructure in key areas becomes more and more important.How to carry out more effective direct monitoring of high-speed rail track equipment and even ordinary track equipment and master the relevant parameters of track state at the same time is the technical content that needs to be studied.Infrastructure temperature is an important factor to determine the stability of seamless line.Using video network data monitoring technology,the on-site video data,infrastructure temperature data and track displacement data are superimposed and displayed in real time,which can provide more convenient and effective equipment monitoring means for operation and maintenance units and has practical application significance.(1)Dual-spectrum infrared thermal imaging thermometry camera was selected for video data acquisition.While the temperature measuring camera collected the surface temperature of the track in real time,real-time measurement of the track displacement scale in the video image was carried out based on the principle of image recognition,and finally the function of the monitoring system was completed.The temperature and displacement data are superimposed in real time and displayed on the monitoring video to realize the scheme design of the track displacement and temperature and video superimposed monitoring system.(2)Aiming at the low accuracy of track temperature measurement camera in field environment,a Improved Fuzzy C-Means Segmentation(IFCM)and Otsu algorithm are proposed to improve track infrared image segmentation and temperature extraction method.The proposed method combines the advantages that IFCM algorithm can converge to the ideal clustering center and Otsu algorithm can determine the threshold adaptively,which improves the noise and inter-class variance of track infrared image,and improves the segmentation accuracy and efficiency.Finally,the Hough linear detection method is used to accurately divide the track image and calibrate the coordinates according to the linear shape of the track,and the track temperature is extracted according to the coordinates,so as to overcome the influence of the environment on the temperature measurement and achieve high precision temperature extraction.(3)The measurement methods of track displacement change are mostly contact measurement and visual recognition of complex devices,which will affect the safety of track and have low working reliability.In view of the above problems,a rail longitudinal displacement identification method based on image processing is proposed,which integrates the advantages of convenient image processing and high reliability of real-time monitoring.It can overcome the influence of all kinds of harsh environment on the image.Firstly,the image quality is improved by improving the Gaussian filtering algorithm,which can remove the noise of the displacement scale image while preserving the details of the image.Secondly,based on the color space segmentation principle of HSV(Hue,Saturation,Value)and the multi-threshold segmentation algorithm of Otsu,backgrounds except yellow displacement ruler and yellow pointer are segmented.Finally,edge detection was introduced to supplement image edge information.The spatial Angle tilt of the image caused by infrared temperature camera vibration or track vibration was corrected based on radon principle,and the displacement data was calculated by distance method to realize the identification and monitoring of track displacement.(4)SDK decoding is used to meet the design requirements of character superposition for monitoring videos,realize the superposition of track temperature and displacement characters,and realize the WEB preview and playback of real-time videos at the same time.In this way,the superposition consumes less energy and reduces the delay of character superposition.The real-time measurement of temperature and displacement for several days is carried out to realize the real-time monitoring of the track condition,which verifies the stability and feasibility of the monitoring system.This thesis has 44 figures,11 tables and 95 references.