Railway wagon loading state detection system based on 2D3D fusion

Abstract

The utility model relates to a railway wagon loading state detection system based on 2D3D fusion belongs to the field of track vehicle three -dimensional imaging, solves the problem that the existing railway vehicle state monitoring system causes the problem that the measurement precision is influenced greatly based on image automatic identification, three multi -line laser radars of the utility model are installed to the top of truss and left and right side surface, and the controller is installed at the bottom of truss, one magnetic steel speed sensor is set in the direction of coming train of truss according to preset distance, and is fixed outside one side rail, and the remaining two magnetic steel speed sensors are set in the direction of leaving train of truss, and are set outside the same side rail, and there is interval between the two, and the controller carries out coordinate transformation, fusion, superposition and splicing according to three multi -line laser radars gathering three -dimensional point cloud data in proper order, and then combines speed information, restores three -dimensional point cloud image of whole railway wagon, extracts the profile size of goods in each compartment in the image, realizes loading state monitoring, and is mainly used for wagon loading state monitoring.

Description

Technical Field

[0001] This utility model relates to the field of three-dimensional imaging technology for rail vehicles. Background Technology

[0002] Radar array scanning is a composite detection method that uses multiple distributed, multi-line, high-precision lidars as the primary detection means, supplemented by speed measuring devices, to generate a 3D image of the target by illuminating the moving train. The controller, acting as the brain of the entire system, is responsible for activating the radar, transmitting and storing speed information, acquiring the 3D point cloud data generated by the radar, fusing and stitching multi-radar data, and ultimately generating a 3D model of the entire train. The study has developed a method for quickly and conveniently acquiring 3D point cloud graphics, measuring spatial coordinates, and identifying and warning of dimensional issues related to vehicles and their cargo during high-speed railway freight train travel.

[0003] Existing railway traffic status monitoring systems mostly employ linear scan cameras with supplementary lighting. While offering high imaging accuracy, high resolution, and good interactive effects, their visible light imaging principle makes them susceptible to the influence of sunlight, including shadows, direct sunlight, reflections, and changes in vehicle and cargo status due to rain, snow, and maintenance. This significantly impacts the measurement accuracy of image-based automatic recognition and easily leads to false alarms. In certain areas, particularly in railway freight car oversize detection and loading status monitoring systems, there is still considerable room for improvement regarding the lack of depth dimensional information and spatial coordinates, as well as operational safety. Therefore, it is necessary to find an automated detection method capable of generating dynamic dimensions and spatial coordinates of railway freight vehicles during transit. This method would enable automatic analysis, identification, and early warning of issues related to dimensions and coordinates during freight inspection operations, such as changes in car body shape, deformation and anomalies of vehicle components, cargo imbalance, cargo displacement, full-end oversize, reinforcement of oversize cargo loading, and dimensional verification during transit. This would assist in high-definition freight inspection identification, improve the accuracy of oversize detection, and enhance the overall level of automated freight inspection. Utility Model Content

[0004] The purpose of this invention is to solve the problem that the measurement accuracy of existing railway traffic status monitoring systems is greatly affected by image-based automatic recognition, and to provide a railway freight car loading status detection system based on 2D-3D fusion.

[0005] The railway freight car loading status detection system based on 2D3D fusion includes 3 multi-line lidars, a truss, 3 magnetic speed sensors and a controller;

[0006] The three multi-line lidar units are installed on the top and left and right sides of the truss, and the controller is installed at the bottom of the truss.

[0007] Of the three magnetic speed sensors, one magnetic speed sensor is set at a preset distance in the oncoming direction of the truss and fixed on the outside of one side rail. The remaining two magnetic speed sensors are set in the departing direction of the truss, and are set on the outside of the same side rail with a gap between them.

[0008] When a railway freight car wheel runs over a magnetic speed sensor, the magnetic speed sensor generates a trigger signal.

[0009] The trigger signal output terminals of the three magnetic speed sensors are all connected to the trigger signal input terminal of the controller. The controller is used to control the three lidars to perform scanning operations based on the first trigger signal output by the magnetic speed sensor in the direction of oncoming traffic; it is also used to control the three lidars to stop working based on the last trigger signal of the magnetic speed sensor farthest from the truss in the direction of departure from the vehicle; and it is also used to obtain vehicle speed information based on the time interval between the trigger signals output by the two magnetic speed sensors in the direction of departure from the vehicle and the distance between the two magnetic speed sensors.

[0010] Three multi-line lidars were used to collect three-dimensional point cloud data of the top surface and left and right sides of the railway freight cars, respectively.

[0011] An array of three multi-line lidar sensors is connected to a controller via a network cable. The controller is used to perform coordinate transformation, fusion, overlay, and stitching on the three-dimensional point cloud data of the top surface and left and right sides of the railway freight cars collected by the three lidar sensors. Combined with the train speed information, the controller reconstructs the three-dimensional point cloud image of the entire railway freight car train and extracts the outline dimensions of the cargo in each car from the three-dimensional point cloud image of the entire railway freight car train, thereby realizing the monitoring of the loading status of the railway freight cars.

[0012] Preferably, the preset distance between a magnetic speed sensor in the direction of oncoming traffic and the truss is 15 to 20 meters.

[0013] Preferably, the two multi-line lidars located on the left and right sides of the truss are installed at a height of 2.5 meters above the ground.

[0014] Preferably, the three multi-line lidar units are installed in the same plane.

[0015] Preferably, the multi-line lidar uses multiple beams of invisible near-infrared laser light emitted simultaneously to acquire three-dimensional point cloud data.

[0016] Preferably, the controller is encapsulated in a waterproof electrical box and installed at the bottom of the truss, on the outside of the two steel rails.

[0017] Advantages of this utility model:

[0018] The railway freight car loading status detection system based on 2D-3D fusion provided by this utility model utilizes multiple distributed multi-line lidars to perform near-infrared laser non-contact measurement on the moving train, scanning the train omnidirectionally without blind spots and generating point cloud data; the system obtains and stores the train speed information by using the distance between two magnetic speed sensors and the time difference between their received signals; the controller performs coordinate transformation, fusion, superposition, and stitching on the measured point cloud data, and supplements it with speed information to generate a complete three-dimensional point cloud image of the railway freight car;

[0019] In practical applications, the system stores the 3D point cloud images of railway freight cars obtained each time, based on the storage instructions received from the user. This enables data backtracking of train digitization, coordinate representation, and dimensional representation. Operators can download the 3D point cloud images by accessing the remote server, which is convenient, fast, and easy to operate. Attached Figure Description

[0020] Figure 1 This is the overall architecture diagram of the railway freight car loading status detection system based on 2D3D fusion described in this utility model;

[0021] Figure 2 It is a diagram showing the relative positions of the railway freight cars and the truss from a side view.

[0022] Figure 3a This is a schematic diagram of the installation of a multi-line lidar mounted on top of a truss.

[0023] Figure 3b This is a schematic diagram of the installation of a multi-line lidar mounted on the left side of the truss.

[0024] Figure 3c This is a schematic diagram of the installation of a multi-line lidar mounted on the right side of the truss. Detailed Implementation

[0025] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0027] Specific Implementation Method 1: Combination Figure 1 and Figure 2This embodiment describes a railway freight car loading status detection system based on 2D-3D fusion, which includes three multi-line lidar units 1, a truss 2, three magnetic speed sensors 3, and a controller.

[0028] The three multi-line lidar units 1 are installed on the top and left and right sides of the truss 2, and the controller is installed at the bottom of the truss 2;

[0029] Of the three magnetic speed sensors 3, one magnetic speed sensor 3 is set at a preset distance in the oncoming direction of the truss 2 and fixed on the outside of one side rail. The remaining two magnetic speed sensors 3 are set in the departing direction of the truss 2, and are set on the outside of the same side rail with a gap between them.

[0030] When a railway freight car wheel runs over the magnetic speed sensor 3, the magnetic speed sensor 3 generates a trigger signal;

[0031] The trigger signal output terminals of the three magnetic speed sensors 3 are all connected to the trigger signal input terminal of the controller. The controller is used to control the three lidars to perform scanning operations based on the first trigger signal output by the magnetic speed sensor 3 in the direction of oncoming vehicle; it is also used to control the three lidars to stop working based on the last trigger signal of the magnetic speed sensor 3 that is farthest from the truss 2 in the direction away from the vehicle; it is also used to obtain vehicle speed information based on the time interval between the trigger signals output by the two magnetic speed sensors 3 in the direction away from the vehicle and the distance between the two magnetic speed sensors 3.

[0032] Three multi-line lidar units 1 respectively collect three-dimensional point cloud data of the top surface and left and right sides of railway freight cars;

[0033] An array of three multi-line lidar sensors is formed and connected to a controller via a network cable. The controller is used to perform coordinate transformation, fusion, superposition, and stitching on the three-dimensional point cloud data of the top surface and left and right sides of the railway freight cars collected by the three lidar sensors. Combined with the train speed information, the three-dimensional point cloud image of the entire railway freight car train is restored. The outline dimensions of the cargo in each car are extracted from the three-dimensional point cloud image of the entire railway freight car train, thereby realizing the monitoring of the loading status of the railway freight cars.

[0034] This embodiment provides a specific structure for a railway freight car loading status detection system based on 2D-3D fusion. The system defines the layout and cooperation relationship between three multi-line lidars 1, a truss 2, and three magnetic speed sensors 3, as well as the area to be collected. The system is combined with a controller to process the data to achieve loading status monitoring. The data processing process of the controller can be implemented using existing technologies. The system has a simple structure and high detection accuracy.

[0035] The working process of this utility model controller is as follows: It continuously monitors the trigger signal sent by the magnetic speed sensor 3. When a passing train is detected, it controls the three multi-line lidars 1 on the left, right, and top surfaces to start laser scanning and data acquisition. During this monitoring process, it simultaneously analyzes and saves the train speed information and receives the point cloud data collected by each multi-line lidar 1 until the passing train is over. After the passing train is over, the single-surface point cloud data from the left, right, and top surfaces are combined into a three-dimensional point cloud image of the entire train of freight cars. At the same time, the three lidars stop scanning and enter a standby state.

[0036] In this utility model, the magnetic steel speed sensor 3 can also be used to count axles and determine the vehicle. The generated signal is stored in the controller and used for later point cloud reconstruction to calculate the length of the train.

[0037] It should be noted that, referring to 3a to 3c, the three multi-line lidars 1 employ a large field-of-view fan-shaped scanning. When installing the top lidar, it should be roughly aligned downwards with the rail of interest. When installing the two side lidars, the installation height of the two multi-line lidars 1 located on the left and right sides of the truss 2 is 2.5 meters above the ground. They are horizontally illuminating the rail of interest, and no leveling device is required for installation; approximate horizontal alignment is sufficient. Later calibration will achieve coordinate fusion of the three lidars. After calibration, as long as the attitude of the three lidars remains stable and does not deflect or move, repeated calibration and adjustment are unnecessary. A composite calibration can be performed periodically every six months to ensure the accuracy of the 3D imaging. The three multi-line lidars 1 should be installed in the same plane as much as possible, and the preset distance between the magnetic speed sensor 3 in the oncoming direction and the truss 2 is 15 to 20 meters. In practical applications, after the three multi-line lidar units 1 are installed, they illuminate and image the railway track separately. By adjusting the coordinate systems of each radar, including coordinate rotation and translation, the railway track images formed by the three radars are made to overlap. The rotation and translation parameters of each radar are then recorded and stored in the controller as radar calibration data. This achieves unification of the three radars at the same coordinate origin, enabling precise fusion and stitching of the data from the three radars. The multi-line lidar units 1 use multiple beams of invisible near-infrared laser light for 3D point cloud data acquisition. The high-speed rotation mechanism and multi-beam firing method inside the three multi-line lidar units 1 ensure the density of point cloud acquisition even at high train speeds. The distributed deployment of the three multi-line lidar units 1 allows for blind-spot-free illumination of the three sides of the railway freight car, obtaining a complete 3D point cloud image of the freight car.

[0038] To effectively protect the controller, it is encapsulated in a waterproof electrical box 4 and installed at the bottom of the truss 2, on the outside of the two steel rails.

[0039] While specific embodiments of the present invention have been described herein with reference to them, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A railway freight car loading status detection system based on 2D / 3D fusion, characterized in that, It includes 3 multi-line lidars (1), a truss (2), 3 magnetic speed sensors (3) and a controller; The three multi-line lidars (1) are installed on the top and left and right sides of the truss (2), and the controller is installed at the bottom of the truss (2). Of the three magnetic speed sensors (3), one magnetic speed sensor (3) is set at a preset distance in the direction of oncoming traffic on the truss (2) and fixed on the outside of one side rail. The remaining two magnetic speed sensors (3) are set in the direction of departure from the truss (2), and the two are set on the outside of the same side rail with a gap between them. When the wheels of a railway freight car run over the magnetic speed sensor (3), the magnetic speed sensor (3) generates a trigger signal; The trigger signal output terminals of the three magnetic speed sensors (3) are all connected to the trigger signal input terminal of the controller; the controller is used to control the three lidars to perform scanning work according to the first trigger signal output by the magnetic speed sensor (3) in the direction of oncoming vehicle; it is also used to control the three lidars to stop working according to the last trigger signal of the magnetic speed sensor (3) that is farthest from the truss (2) in the direction away from the vehicle; it is also used to obtain vehicle speed information according to the time interval between the trigger signals output by the two magnetic speed sensors (3) in the direction away from the vehicle and the distance between the two magnetic speed sensors (3); Three multi-line lidars (1) respectively collect three-dimensional point cloud data of the top surface and left and right sides of railway freight cars; An array of three multi-line lidars (1) is formed and connected to the controller via a network cable. The controller is used to perform coordinate transformation, fusion, superposition and splicing on the three-dimensional point cloud data of the top surface and left and right sides of the railway freight car collected by the three lidar sensors. Combined with the train speed information, the three-dimensional point cloud image of the entire railway freight car is restored. The outline size of the cargo in each car is extracted from the three-dimensional point cloud image of the entire railway freight car, so as to realize the monitoring of the loading status of the railway freight car.

2. The railway freight car loading status detection system based on 2D / 3D fusion as described in claim 1, characterized in that, The preset distance between a magnetic speed sensor (3) in the direction of oncoming traffic and the truss (2) is 15 to 20 meters.

3. The railway freight car loading status detection system based on 2D / 3D fusion according to claim 1, characterized in that, The two multi-line lidars (1) located on the left and right sides of the truss (2) are installed at a height of 2.5 meters above the ground.

4. The railway freight car loading status detection system based on 2D / 3D fusion according to claim 1, characterized in that, Three multi-line lidars (1) are installed in the same plane.

5. The railway freight car loading status detection system based on 2D / 3D fusion according to claim 1, characterized in that, Multi-line lidar (1) uses multiple beams of invisible near-infrared lasers emitted simultaneously to acquire three-dimensional point cloud data.

6. The railway freight car loading status detection system based on 2D / 3D fusion according to claim 1, characterized in that, The controller is encapsulated in a waterproof electrical box (4) and is placed at the bottom of the truss (2) on the outside of the two steel rails.

Images