Road vehicle type identification system

Abstract

A road vehicle type recognition system comprises a central control calculation unit and a plurality of side shooting cameras arranged above a road, target view fields of the side shooting cameras cover all lanes in the same direction, and the number of the side shooting cameras is smaller than that of the lanes; the side shooting camera is configured to obtain a lane section image or a video stream in a target view field and send the lane section image or the video stream to the central control calculation unit; and the central control calculation unit is configured to perform vehicle type identification according to the received lane section image or video stream. Compared with a traditional scheme, the number of side shooting cameras can be reduced, the system cost is saved, and the data transmission pressure is reduced.

Description

Technical Field

[0001] The utility model relates to the field of intelligent transportation, and particularly to a road vehicle type recognition system. Background Art

[0002] The vehicle type recognition technology has a wide range of applications in the field of intelligent transportation. For example, based on vehicle type recognition, the road toll system can adjust the toll standard according to the vehicle type; the vehicle type recognition technology can be integrated with the electronic toll collection system to achieve automatic toll collection without manual intervention; through the vehicle type recognition technology, it is possible to monitor whether a vehicle is driving on a section where tolls should be paid without payment, or to achieve the purpose of evading tolls by changing the vehicle type in the ETC card, so as to monitor and punish road violation behaviors; the vehicle type recognition technology can also be used for road traffic flow statistics to provide data support for traffic planning and road management, and so on.

[0003] Currently, the mainstream vehicle type recognition systems mostly adopt a scheme of setting one camera for each lane, that is, the number of cameras is equal to the number of main lanes in the section. After obtaining the vehicle images of the lane by using the camera, vehicle type recognition is carried out. On the one hand, this layout method has too high hardware costs. On the other hand, a large number of cameras also bring huge pressure to the data stream transmission of the system.

[0004] Therefore, it is necessary to provide a new road vehicle type recognition system to save system costs and reduce data transmission pressure. Summary of the Utility Model

[0005] This patent provides a road vehicle type recognition system. Compared with the traditional scheme, it can reduce the number of side cameras, save system costs and reduce data transmission pressure. Specifically, it includes:

[0006] A road vehicle type recognition system, comprising:

[0007] A central control computing unit and a plurality of side cameras arranged above the road, the target fields of view of the plurality of side cameras cover all lanes in the same direction. Among them, the number of the side cameras is less than the number of lanes, and the plurality of side cameras are all communicatively connected to the central control computing unit;

[0008] The side cameras are configured to obtain lane cross-section images or video streams within the target fields of view and send them to the central control computing unit;

[0009] The central control computing unit is configured to perform vehicle type recognition according to the lane cross-section images or video streams.

[0010] Optionally, the vertical distance between the lower edge of the longitudinal field of view of the side camera and the nearest lane line in its corresponding target field of view is greater than or equal to 1 / 16 of the longitudinal field of view; and,

[0011] The vertical distance between the upper edge of the longitudinal field of view of the side shooting camera and the nearest lane line in its corresponding target field of view is greater than or equal to 1 / 6 of the longitudinal field of view;

[0012] The field of view of the side shooting camera covers the target field of view of the side shooting camera.

[0013] Optionally, the installation height of the side shooting camera from the ground is 3m - 8m.

[0014] Optionally, the included angle between the side shooting camera and the road driving direction is 70° - 110°.

[0015] Optionally, the included angle between the side shooting camera and the road driving direction is 90°.

[0016] Optionally, the included angle between the side shooting camera and the horizontal plane is configured to be adjustable, and by adjusting the included angle between the side shooting camera and the horizontal plane, the target field of view of the side shooting camera is adjusted.

[0017] By arranging multiple side shooting cameras above the road, on the basis that the number of side shooting cameras is less than the number of lanes, the target fields of view of all side shooting cameras cover all lanes in the same direction, that is, the target field of view of at least one side shooting camera covers two lanes. Through this arrangement method, the number of side shooting cameras used can be reduced, thereby saving the cost of the system. On the basis of reducing the number of side shooting cameras used, the amount of data transmitted by the side shooting cameras to the central control computing unit is reduced, reducing the pressure of data transmission. Description of the Drawings

[0018] In order to make the content of the present utility model easier to be clearly understood, the following further details the present utility model according to specific embodiments in conjunction with the drawings, where

[0019] Figure 1 is the system block diagram of the road vehicle type recognition system of the present utility model;

[0020] Figure 2 is the schematic diagram of the longitudinal field of view of the side shooting camera of the present utility model;

[0021] Figure 3 is the schematic diagram of the arrangement of side shooting cameras of a road vehicle type recognition system of the present utility model;

[0022] Figure 4 is the schematic diagram of the arrangement of side shooting cameras of another road vehicle type recognition system of the present utility model. Detailed Embodiments

[0023] The following further details the technical solutions of the present utility model through the drawings and embodiments.

[0024] The road vehicle type recognition system has been widely used in the field of intelligent transportation. For example:

[0025] Dynamic tolling: Based on vehicle type recognition, the road tolling system can adjust the toll standard according to the vehicle type. For example, for large trucks or buses, higher tolls are charged to reflect their greater wear on the road infrastructure.

[0026] Automatic tolling: Vehicle type recognition technology can be integrated with the electronic toll collection system to achieve automatic tolling without manual intervention. When a vehicle passes through the toll station, the system can automatically identify the vehicle type and deduct the corresponding toll accordingly, improving the toll collection efficiency and accuracy.

[0027] Violation monitoring: Through vehicle type recognition technology, the system can monitor whether a vehicle is driving on a section where tolls should be paid without payment, or achieving the purpose of evading tolls by changing the vehicle type in the ETC card, so as to monitor and punish road violation behaviors.

[0028] Traffic flow statistics: Vehicle type recognition technology can also be used for road traffic flow statistics, helping traffic management departments understand the traffic flow of different vehicle types at specific times and locations, providing data support for traffic planning and road management.

[0029] In summary, the application of vehicle type recognition technology in the field of road tolling can improve the toll collection efficiency, accuracy and supervision ability, and contribute to the improvement of road traffic management and service quality.

[0030] Combined Figure 1 As shown, in this embodiment, a road vehicle type recognition system is provided, including a central control computing unit S1 and a plurality of side shooting cameras 1 to N installed above the road. The side shooting cameras can be visual cameras that can stably output video streams or image data frames. Specifically, a plurality of side shooting cameras can be installed on the gantry of the road section or on the roadside L-shaped vertical poles.

[0031] To ensure that vehicle type recognition can be achieved for all vehicles in all lanes, the target fields of view of the plurality of side shooting cameras need to cover all lanes 1 to N+M in the same direction, M≥1. Here, it should be noted that there is usually an emergency lane on the highway, and it can be determined whether the target field of view needs to cover the emergency lane according to different needs. For example, when the vehicle type recognition system is used to integrate with the electronic toll collection system to achieve road tolling, the target field of view can only cover the main lanes (excluding the emergency lane); while when conducting violation monitoring or road traffic flow statistics, the target field of view can cover the emergency lane. For example, during holidays, the traffic flow on the highway is large, and some vehicles illegally occupy the emergency lane to drive. At this time, it is necessary to identify the vehicles in the emergency lane for subsequent violation handling by the traffic management department, etc. No specific scenarios are limited here.

[0032] Among them, to achieve the purpose of the present utility model, the number of side cameras is less than the number of lanes. That is, among the multiple installed side cameras, the target field of view of at least one side camera covers at least two adjacent lanes (for example Figure 1 in the side camera 2, the target field of view covers lane 2 and lane 3), the target field of view of the side camera is the target area of concern in the field of view of this side camera. By this method, the number of side cameras used can be reduced, that is, the number of side cameras used can be less than the number of lanes in the road section, so as to save the system hardware cost. At the same time, because the number of side cameras used is reduced, the amount of image data uploaded by this side camera can be reduced, thus reducing the data transmission pressure between the side camera and the central control computing unit.

[0033] The multiple side cameras are all communicatively connected to the central control computing unit, and the connection method can be wired and / or wireless; the side camera is configured to acquire the road section image or video stream within the target field of view and send it to the central control computing unit; the central control computing unit is configured to perform vehicle type recognition according to the road section image or video stream. The side camera can acquire a video stream or take continuous photos at a preset frequency or be triggered to take photos by a signal. Among them, the trigger signal can be other devices such as lidar and inductive loop that trigger the camera to take photos when a vehicle is sensed passing by, or the camera itself has a vehicle recognition function, so that when a vehicle passes through its target field of view, it actively triggers to take photos. The side camera sends the acquired video stream or image to the central control computing unit for recognition processing. The central control computing unit can be a computer, industrial control computer, server, etc. that at least has the function of receiving and processing video and image data. The central control computing unit can integrate some mature image algorithms in the prior art for vehicle type recognition, such as feature-based algorithms, deep learning-based algorithms, support vector machine algorithms, etc., which are not specifically limited here.

[0034] In the above embodiment, by the method that the target field of view of at least one side camera covers at least two lanes, the number of side cameras used can be effectively reduced, and then the effects of saving system cost and reducing data transmission pressure can be achieved.

[0035] In one embodiment, as Figure 2As shown in the figure, it is a schematic diagram of the longitudinal field of view of the side camera, and the field of view of the side camera covers the target field of view of the side camera. In the figure, the target field of view of the side camera includes target lane 1 and target lane 2. Target lane 1 is close to the side camera, and target lane 2 is far away from the side camera. In order to ensure that the complete outline of the vehicle in the target lane can be obtained and the vehicle outline is as close to the center of the field of view as possible, the vertical distance between the lower edge of the longitudinal field of view of the side camera and the nearest lane line in the corresponding target field of view (corresponding to the target lane closest to the side camera in the figure, that is, the lower lane line of target lane 1) is greater than or equal to 1 / 16 of the longitudinal field of view; the vertical distance between the upper edge of the longitudinal field of view of the side camera and the nearest lane line in the corresponding target field of view (corresponding to the upper lane line of the farthest target lane in the figure, that is, target lane 2) is greater than or equal to 1 / 6 of the longitudinal field of view.

[0036] In this embodiment, by configuring the field of view coverage area of ​​the side camera, it can be effectively ensured that the side camera can obtain the complete outline of the vehicle in the target field of view and place the vehicle in the center area of ​​the field of view, effectively avoiding the target vehicle at the edge of the field of view. The distortion at the edge of the image is large, which leads to a decrease in vehicle model recognition rate.

[0037] In one embodiment, the side camera is installed at a height of 3m-8m from the ground. As mentioned above, the side camera can be installed on a gantry of a road section or an L-shaped pole on the side of the road. The installation height varies according to different scenarios, but it is necessary to ensure that the side camera can obtain a complete vehicle image in the target field of view at the installation height.

[0038] In one embodiment, the angle between the side camera and the direction of travel on the road is 70°-110°, for example, 70°, 75°, 80°, 85°, 90°, 95°, 100°, 105°, 110°, etc. can be set according to actual conditions, and no specific limitation is made here. In actual applications, the angle between the side camera and the direction of travel on the road will affect the quality of the image captured by the camera. When the angle is too large or too small, the captured image will have a large distortion, which is not conducive to subsequent image processing and affects the final vehicle model recognition result. However, in some installation environments, the mounting structure of the side camera will interfere with the side camera. For example, the door frame or the pole of the road itself may affect the field of view of the camera. In this case, the side camera needs to deviate from the mounting structure at a certain angle. In actual applications, the deviation angle is within 20°, which can ensure that it is not blocked by the mounting structure and the acquired image will not have a large distortion.

[0039] In one embodiment, when the mounting structure of the side camera does not interfere with the field of view of the side camera, the angle between the side camera and the driving direction of the road can be set to 90°. At this time, the vehicle image obtained by the side camera is minimally distorted, which facilitates subsequent image processing and vehicle model recognition.

[0040] In one embodiment, the included angle between the side shooting camera and the horizontal plane is configured to be adjustable, that is, the side shooting camera is installed at an adjustable angle. For example, the installation structure may include components such as a bearing structure or a rotating shaft structure that can adjust the angle. By adjusting the included angle between the side shooting camera and the horizontal plane, the target field of view of the side shooting camera can be adjusted. For example, when a certain side shooting camera fails, other cameras can take over the failed camera to obtain the images or video streams within its target field of view. Or when the installation angle of the side shooting camera changes due to long-term use / road vibration and other reasons, resulting in a non-optimal shooting angle, the angle can be adjusted for calibration. The specific angle adjustment method can be manual adjustment by on-site staff or remote adjustment, which is not specifically limited here.

[0041] To more clearly illustrate the technical solution of the present utility model, two specific scenarios are described as follows:

[0042] Scenario 1: As shown in Figure 1 and Figure 3 , a road vehicle type recognition system is provided. Figure 3 The system consists of side shooting camera 1, side shooting camera 2, side shooting camera 3 and a central control computing unit S1. The three side shooting cameras are respectively communicatively connected to S1.

[0043] The system is deployed on a 4-lane gantry. In this scenario, the emergency lane is not used as the target field of view. As Figure 3 is a cross-sectional view of the lane perpendicular to the driving direction. The gantry vertical pole is 3.5 meters high. The three side shooting cameras are arranged on the gantry crossbar as shown in the figure. Side shooting camera 1 is arranged above the left side of lane 1, and its target field of view is lane 2; side shooting camera 2 is arranged above the right side of lane 2, and its target field of view is lane 1; side shooting camera 3 is arranged above the right side of lane 4, and its target field of view is lanes 3 and 4.

[0044] When a vehicle passes through the section, the corresponding side shooting camera will capture its side, and then transmit the captured image to the central control computing unit S1. S1 uses image algorithms for vehicle type recognition, and finally obtains the vehicle type.

[0045] Scenario 2: As shown in Figure 1 and Figure 4 , Figure 4 The system consists of side shooting camera 1, side shooting camera 2, side shooting camera 3 and a central control computing unit S1. The three side shooting cameras are respectively communicatively connected to S1.

[0046] The system is deployed on a 5-lane gantry. In this scenario, the emergency lane is not used as the target field of view. As Figure 4It is a sectional view of the lane perpendicular to the driving direction. The height of the gantry vertical pole is 7.5 meters. Three side cameras are arranged on the gantry crossbar as shown in the figure. Side camera 1 is arranged above the left side of lane 1, and its target field of view is lanes 1 and 2; Side camera 2 is arranged above the left side of lane 2, and its target field of view is lane 3; Side camera 3 is arranged above the right side of lane 5, and its target field of view is lanes 4 and 5.

[0047] When a vehicle passes through the section, the corresponding side camera will capture its side, and then transmit the captured image to the central control computing unit S1. S1 uses image algorithms for vehicle type recognition and finally obtains the vehicle type.

[0048] The side cameras and the central control computing unit involved in the present utility model are well-known technical products in the field, and their specific structures will not be described here. Moreover, the data processing related technologies for the camera to complete vehicle capture and the central control computing unit to complete vehicle type recognition are well-known technologies in the industry.

[0049] Obviously, the above embodiments are only examples for clearly explaining the present utility model, rather than limitations on the implementation modes of the present utility model. For those of ordinary skill in the art, other different forms of changes or variations can be made on the basis of the above description. It is not necessary and impossible to enumerate all implementation modes here. And these obvious changes or variations derived from the spirit of the present utility model are still within the protection scope of the present utility model.

Claims

1. A road vehicle type recognition system, characterized in that: include: A central control computing unit and a plurality of side cameras arranged above the road, wherein the target fields of view of the plurality of side cameras cover all lanes in the same direction, wherein the number of the side cameras is less than the number of lanes, and the plurality of side cameras are all in communication connection with the central control computing unit; The side camera is configured to obtain a lane section image or video stream within a target field of view and send it to the central control computing unit; The central control computing unit is configured to perform vehicle type recognition based on the lane section image or video stream.

2. The road vehicle type recognition system according to claim 1, characterized in that: The vertical distance between the lower edge of the longitudinal field of view of the side camera and the nearest lane line in the corresponding target field of view is greater than or equal to 1 / 16 of the longitudinal field of view; and The vertical distance between the upper edge of the longitudinal field of view of the side camera and the nearest lane line in its corresponding target field of view is greater than or equal to 1 / 6 of the longitudinal field of view.

3. The road vehicle type recognition system according to claim 1, characterized in that: The side-shooting camera is installed at a height of 3m-8m from the ground.

4. The road vehicle type recognition system according to any one of claims 1 to 3, characterized in that: The angle between the side camera and the driving direction of the road is 70°-110°.

5. The road vehicle type recognition system according to claim 1, characterized in that: The angle between the side camera and the horizontal plane is configured to be adjustable, and the target field of view of the side camera is adjusted by adjusting the angle between the side camera and the horizontal plane.

Images