Design of Feature-based Underbody Image Registration Method in TFDS

As the lifeline of the national economy, railway traffic is one of important mode of transportation, it runs fast, has a large load, is safe and is not affected by weather, the rapid development of our country current implementing sustainable development strategy of critical face railway leap-forward development of vehicle safety work put forward new requirements, rail safety technical innovation planning arises at the historic moment. The dynamic image detection system of freight car running fault (TFDS) uses the real-time transmission and processing technology of mass image data of high-speed camera and high-precision positioning technology, and installs high-speed industrial camera beside the track to dynamically identify hidden faults and common faults of freight car running on the line Based on the above background, this essay will conduct an in-depth study on the intelligent image recognition method of railway operation safety and its application in the railway truck operation safety detection and monitoring system. At present, TFDS system only completes the function of image acquisition and transmission, while the automatic recognition module for image anomalies still needs to be solved urgently This project is to use digital image processing and other technologies to design the corresponding abnormal distortion image recognition algorithm, and use the algorithm to complete the correction of the distortion Figure of a train in a short time after the train passes, so as to improve the current TFDS system, conform to and promote the development of railway truck transportation safety.