Intelligent Video Surveillance and Early Alarms Method for Railway Tunnel Collapse

Looking at the problem of railway tunnel entrance collapse, this paper proposes a warning system based on intelligent video surveillance. In the image processing algorithm, HSV color space integrating RGB time domain segmentation combined with double threshold image segmentation technique is used to extract the moving object in the monitoring range. The Sobel edge detection operator and the improved Hough transform algorithm are used to extract the characteristics of the rails to determine the size and position of moving objects. Self-adaptive background update mechanism based on Kalman filter theory eliminates the influence of natural conditions such as weather and light. Feature recognition of collapse is realized based on the image processing algorithm. The slope displacement monitor is introduced to achieve real-time early-warning before collapsing. Simulation experiments show that this method can extract the target features accurately, achieve immediate and effective early warning, and cope with the environmental disturbance factors.