A real-time measurement method for non-contact relative displacement of steel sleepers based on deep learning and perspective transformation

Abstract

The invention discloses a non-contact real-time measurement method for the relative displacement of a steel rail sleeper based on deep learning and perspective transformation, and belongs to the field of high-speed rail track structure monitoring. Paste a signboard containing at least two circles on the sleeper, and at least a pair of sleepers with fixed relative positions are included in the detection range of each detection point; the calibration image is used to train the deep learning model; the initial standard image is collected to calculate the perspective transformation matrix; For the real-time image obtained at each detection point, the original pixel coordinates of the center of the circle are detected through the deep learning model, the pixel coordinates of the center of the circle are converted into the pixel coordinates of the center of the circle after perspective by using the perspective transformation matrix, and the distance between adjacent sleepers in the perspective image is calculated. Combined with the pixel distance between a pair of sleepers with a fixed relative position and the actual distance, the actual distance between adjacent sleepers is obtained as the rough calculation result; then the rough calculation result is smoothed and filtered to obtain the real-time relative displacement The precise value, high measurement accuracy.

Description

technical field [0001] The invention belongs to the field of high-speed rail track structure monitoring, and more specifically relates to a non-contact real-time measurement method for the relative displacement of a steel rail sleeper based on deep learning and perspective transformation. Background technique [0002] In order to meet the needs of various complex regional environments, long continuous girder bridges will inevitably appear on high-speed railways. During the use of the rail telescopic adjuster and the rail lifting device at the beam joints on the relevant lines, there were problems such as tilting and cracking of the sleeper, large deformation or even jamming of the scissors and forks of the rail lifting device, resulting in a large maintenance and repair workload and economic loss. Due to the large maintenance and repair workload of the rail telescopic adjuster, it is one of the three weak links of the high-speed rail track structure. Therefore, the high-spe...

AI Technical Summary

Problems solved by technology

[0004] (1) Generally, it can only be used as a point-of-work monitoring, and the monitoring range in space and time is limited, and large-scale real-time measurement cannot be realized;

[0005] (2) All are contact sensing methods, and the sensors installed on the track structure are a major safety hazard for the operation of high-speed trains

[0007] However, due to the complex environment of the high-speed rail site, various factors such as light, wind and rain, and train vibration will affect the imaging results. In addition to the perspective distortion of the camera itself, the relative displacement of the steel sleeper measured based on a simple image recognition method cannot meet the accuracy of track structure monitoring. Require

Images