Train loading state high-definition monitoring and overloading detection method

Abstract

The invention relates to a train loading state high-definition monitoring and overloading detection method. The method comprises the following steps that a high-definition camera set is used for carrying out real-time monitoring, when the arrival of a train is sensed, the system automatically starts a video system, a linear array CCD (charge-coupled device) camera carries out real-time scanning, an overloading object is analyzed and alarmed, a binocular visual three-dimensional measuring camera is triggered to carry out binocular visual analysis on obtained train images, feature extraction, image matching and train limit contour comparison, then the overloading size and overloading class are obtained, meanwhile, an image analysis technique is adopted to intelligently identify the opening and closing of a box card door, the damage of a canvas and the deviation of cargos, and the abnormal conditions are alarmed and are stored into high-definition pictures. The method has the advantage that the train is monitored by the high-definition videos in the real-time way, the linear array CCD camera is combined to carry out scanning and detection, and the train contour is measured by the binocular visual three-dimensional measuring camera in the three-dimensional way, so the three-dimensional coordinates of the overloading object are obtained, the real-time overloading alarming is realized, and the overloading data and overloading class with accuracy reaching millimeter level can be accurately measured.

Description

technical field [0001] The invention relates to a high-definition monitoring and overrun detection method of a train loading state, belonging to the field of railway safety monitoring. Background technique [0002] In the process of railway cargo transportation, due to the influence of many factors such as line conditions, vehicle conditions and traction speed, the cargo and reinforcement materials are often displaced, resulting in potential safety hazards. In recent years, accidents caused by overloading of trucks have occurred from time to time. [0003] According to relevant statistics, my country's railway department suffers compensation of up to 150 million yuan every year due to the theft of railway goods; railway police often need to play back surveillance videos to find clues. Some criminals fled everywhere by hiding in the gondola compartment, avoiding search. How to effectively monitor these vehicles is also a problem that needs to be solved at present. [0004]...

AI Technical Summary

Problems solved by technology

Infrared monitoring technology is simple and easy, but there are obvious defects: it can only be determined that there is a level of overrun in a certain direction, and the specific overrun position and overrun level cannot be determined; it cannot fully detect the overrun of the goods, and the lower overrun cannot be detected. One of the overrunning pair pipes in the middle is on the ground, which is easily covered by dust and dirt, resulting in a large number of false alarms; infrared laser measurement cannot save the on-site detection situation in the form of pictures, at best it can only record the moment of overrunning, but according to this It is impossible to determine the position and size of the out-of-limit objects at all, and it is impossible to conduct a review. At the same time, it is difficult to distinguish the field disturbances (such as flying insects, paper, etc.) from the actual out-of-limit objects

But there are also obvious defects: the scanning frequency of the most advanced laser radar is 100HZ, that is, the scanning time of the radar is 10 milliseconds, assuming that the train travels at a speed of 36km / h, 10mm per millisecond, that is to say, the detection of the laser radar The sensitivity is 100mm, that is, objects that are less than 100mm out of limit, the laser radar may not be able to detect them

This method can further measure the size of the overrun object beyond the first-level limit, and roughly estimate the over-limit level, but this method also has obvious shortcomings: the obtained size exceeds the first-level limit size, and the over-limit level cannot be really obtained. Overrunning background panels are installed on the ground on both sides of the track, and are easily polluted by objects dropped by the train, which affects image analysis and produces a large number of false alarms, sometimes thousands of false alarms, that is, false alarms for the entire train , it cannot be used at all, because it is a two-dimensional plane measurement, and the goods below 1250mm cannot be obtained beyond the limit

If it is judged to be exceeding the limit, it may be a false alarm caused by the pollution of the background plate. If it is judged not to exceed the limit, it may also be that the exceeding limit is less than 100mm, and the laser radar has not detected it

Therefore, this comprehensive detection cannot fundamentally solve the problem

Images