WO2021077766A1

WO2021077766A1 - Large-area multi-target traffic event detection system and method

Info

Publication number: WO2021077766A1
Application number: PCT/CN2020/095556
Authority: WO
Inventors: 张军; 章庆; 许古午
Original assignee: 南京慧尔视智能科技有限公司
Priority date: 2019-10-24
Filing date: 2020-06-11
Publication date: 2021-04-29
Also published as: CN110718062B; CN110718062A

Abstract

A large-area multi-target traffic event detection system and method. Said method comprises: using a microwave detector to detect, in real time, various target information on a road, the target information comprising the position, speed and driving lane; the microwave detector sending a target position to a video detector, and the video detector detecting feature information of targets; a traffic analyzer matching the information detected by the microwave detector with the information detected by the video detector, to acquire the position, speed, vehicle model and driving lane corresponding to each target; and the traffic analyzer comparing the driving lane, vehicle model, and driving speed of each target with a set speed limit value of each vehicle model within each lane or with a specified vehicle model within each lane, to determine whether a traffic event has occurred. Combining microwave and video is beneficial for a traffic law enforcement department to effectively manage and control vehicle speeding behavior within a speed limit area and illegal lane occupation.

Description

Large-area multi-target traffic incident detection system and method

Technical field

The invention discloses a large-area multi-target traffic incident detection system and method, and relates to the technical field of traffic incident detection.

Background technique

When analyzing the causes of traffic accidents, it is found that speeding is one of the main causes of road traffic accidents. Speeding not only leads to an increase in the number of traffic accidents, but also aggravates the severity of the accident. Although most road sections have speed limit warning signs, drivers are often prone to speeding when there are not many road vehicles, which can lead to serious accidents. Therefore, law enforcement is the key to complying with traffic speed limits, and it is also the last line of defense to curb speeding violations. Only when drivers who do not comply with laws and regulations are severely punished can drivers drive in accordance with regulations and ensure road safety. On many roads, the public security department has already put into use scientific equipment, using different forms of speeding law enforcement equipment such as radar speedometers, laser speedometers, electronic monitoring, etc., to improve the level of road traffic management.

Most of the existing speeding monitoring systems use narrow-beam radar speedometers and electronic monitoring systems to record speeding behavior. The main method is to drive the camera to capture after detecting a speeding vehicle at a fixed detection line, and then perform high-speed image processing and recognition Technology to send the photos taken back to the command center. However, this system is only suitable for monitoring speeding events in a single lane and at a fixed location. It is necessary to install a speed measuring radar in each lane. Moreover, due to the problem of multiple radar coverage areas, it is difficult to achieve one-to-one correspondence between detection targets, and leakage is likely to occur. The phenomenon of multiple inspections and multiple inspections, and it is easy to cause accidents such as rear-end collisions due to rapid speed reduction and rapid acceleration of vehicles.

Summary of the invention

Aiming at the defects in the above-mentioned background technology, the present invention provides a large-area multi-target traffic incident detection system and method, avoiding missed inspections and multiple inspections, and realizing dynamic monitoring of traffic incidents for all targets in the area.

In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is as follows: a large-area multi-target traffic incident detection system, including a microwave detector, a video detector, a traffic analyzer, a switch, and an event platform;

RS485 serial communication mode is used for data transmission between the microwave detector and the video detector;

The microwave detector and the traffic analyzer, the video detector and the traffic analyzer are all interconnected through the switch, and the RJ45 network communication method is used for two-way data transmission;

The traffic analyzer and event platform use network communication for data transmission;

The microwave detector collects target information in the monitoring area and sends the target information to the traffic analyzer; the video detector records road video information in the monitored area in real time, and sends the collected video information to the traffic analyzer; and When the target enters the monitoring area, the microwave detector sends a trigger signal to the video detector, and the video detector receives the trigger signal to capture the target image, identify the captured image feature information, and send the image with the feature information to the traffic analysis Instrument; The traffic analyzer receives target information, video information and images with characteristic information, performs information matching of the same target, and generates complete target information;

At the same time, the traffic analyzer receives the target information of the microwave detector in real time, and determines the traffic incident according to the preset event detection threshold; if the target is determined as a traffic incident, the microwave detector sends a trigger signal to the video detector, and the video detector responds The target captures and sends the combined image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target's traffic incident and sends it to the event platform together with the target information and images detected by the detector, and the event platform finally pushes the traffic event alarm information; if The target is not judged as a traffic incident, the traffic analyzer saves the road video information of the monitored area, and does not report to the event platform; the event platform displays the traffic event related information after receiving the video information, target information and images from the traffic analyzer .

Further, the target information includes target motion information and a target driving lane, the target motion information includes the speed and position of the target, and the characteristic information includes license plate and vehicle type.

Further, when the microwave detector detects that a target enters the monitoring area, it sends a trigger signal to the video detector through the RS485 interface. The video detector receives the trigger signal, and the video detector will identify the feature information of the captured image. And send the image recognition result to the traffic analyzer.

Further, when the microwave detector detects that the target leaves the monitoring area, it sends a trigger signal to the video detector through the RS485 interface, and the video detector receives the trigger signal, captures the target image and sends the image to the traffic analyzer.

Further, the trigger signal includes: the position of the target and a timestamp; the video detector directly captures and recognizes the image of the target according to the target position; the timestamp is the time when the target enters the monitoring area, the time when it leaves the monitoring area, and The time of the traffic incident.

A method for detecting multi-target traffic incidents in a large area includes the following steps:

S1: Deploy microwave detectors and video detectors on the rods, adjust the time of the two through a traffic analyzer, and set the event detection threshold;

S2: The microwave detector tracks the target information on the road in real time and sends the target information to the traffic analyzer; the video detector records the road video information of the monitored area in real time, and the video detector receives the trigger signal of the microwave detector and Identify the image information and send the identified image information to the traffic analyzer;

S3: The traffic analyzer receives the target information tracked by the microwave detector in real time, and determines the traffic incident based on the preset event detection threshold. The traffic analyzer compares the complete information of each target with the set event detection threshold to determine whether traffic occurs Event, when the traffic incident is not established, go to S4; when the traffic incident is established, go to S5;

S4: The traffic analyzer saves the road video information of the monitored area, and does not report to the event platform;

S5: The microwave detector sends a trigger signal to the video detector, the video detector captures the target, and sends the captured image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target where the traffic incident occurred and detects it together with the detector Send the target information and images to the event platform;

S6: The event platform pushes traffic event alarm information.

Further, the complete information includes target information and characteristic information, the target information includes target motion information and target driving lanes, the target motion information includes the speed and position of the target, and the characteristic information includes license plate and vehicle type. ; When the microwave detector detects that a target enters the monitoring area, the microwave detector sends a trigger signal to the video detector, the video detector captures the target, and recognizes the feature information of the image, and the result of the image recognition is sent to the traffic analyzer;

When the microwave detector detects that the target leaves the monitoring area, the microwave detector sends a trigger signal to the video detector, and the video detector receives the trigger signal, captures an image of the target and sends the image to the traffic analyzer.

Further, the traffic incidents include speeding incidents and road occupation incidents.

Further, in S5, when it is judged that a traffic incident occurs, the target is tracked in real time until the target stops the traffic incident or leaves the monitoring area, and then generates three time records, the first is the time t ₁ when the target enters the detection area, and the second is _{Is the time t 2} when the target is detected for the first time, and the third is the time t _{3 when the} target leaves the monitoring area; then the video from t ₂ to t ₃ _{and t 1} , t ₂ , t _{3 are intercepted} Pictures of these three time nodes; finally report the video and 3 pictures of the traffic incident.

Beneficial effects:

1: The present invention changes the existing detection method of taking photos to obtain evidence at a certain fixed detection line. It can be achieved by adopting microwave detectors to detect the speed of multiple targets in multiple lanes in a large area in real time and video detectors to capture evidence. One-to-one correspondence between the detection target and the event record in the detection area, realizing the detection of speeding events at any position in multiple lanes in a large area, effectively avoiding missed or multiple inspections;

2: The present invention is composed of a single microwave detector, video detector, and traffic analyzer, which is convenient for debugging and maintenance;

3: The invention adds the determination of the vehicle type, regulates the vehicle type and speed of each lane on the expressway, and facilitates vehicle management and control.

Description of the drawings

Figure 1 is a block diagram of the system of the present invention;

Figure 2 is a schematic flow chart of the method of the present invention;

Figure 3 is a schematic diagram of the recognition results of each target in the image.

Detailed ways

The implementation of the technical solution will be described in further detail below in conjunction with the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present invention more clearly, and cannot be used to limit the protection scope of the present invention.

An embodiment provided by the present invention: as shown in Fig. 1, a large-area multi-target traffic incident detection system includes a microwave detector, a video detector, a traffic analyzer, a switch, and an event platform;

The traffic analyzer and platform use network communication for data transmission;

Microwave detector:

The microwave detector is responsible for tracking and detecting targets (especially motor vehicles) in the monitoring area. The specific tracking process is:

The transmitting end of the microwave detector emits a microwave signal. The microwave signal meets the target and reflects the echo signal. The microwave detector extracts the distance, angle, speed and other information of the target relative to the microwave detector from the echo signal, which can be based on the distance and angle. Calculate the (x, y) position of the target. The specific extraction of information such as distance, angle, and speed from the echo signal belongs to the conventional technology of microwave detection in the prior art. For details, please refer to the prior art, which will not be repeated here.

After setting the monitoring area of the microwave detector, as long as it is detected that the target reaches the boundary of the monitoring area for the first time (the distance from the microwave detector reaches the specified value), it is considered that the target is detected to enter the monitoring area. When detecting that the target enters the monitoring area, the microwave detector correlates each frame of information of the target in the monitoring area to form the target's motion trajectory to achieve tracking.

When it detects that the target enters the monitoring area, it sends a trigger signal to the video detector to capture the target, and sends the target motion information (such as the target's speed and location) to the traffic analyzer.

Video detector:

The video detector is responsible for image acquisition of the target on the road in the monitoring area, and after receiving the trigger signal of the microwave detector, it captures the target and performs image processing to obtain characteristic information, including identifying the model of the target vehicle and identifying the license plate number (recognition The license plate number is used to report the speeding incident later to know which car it is.) And send the collected video stream information and feature results to the traffic analyzer. Most of the existing video detectors have the functions of video stream collection and license plate recognition. In this application, it is only necessary to use the existing video detectors on the market that can realize this function.

The detectable area of the microwave detector and the video detector can be large or small according to their own design. In practical applications, since these two devices need to work at the same time, take the smaller area that can be detected by the two devices. For example, how many meters in a few lanes. Or, according to actual needs, a range that needs to be monitored is designated as the monitoring area.

Traffic analyzer:

The traffic analyzer is responsible for receiving the target motion data of the microwave detector and the feature results of the video detector, performing target matching and data fusion, generating complete target information, and determining traffic incidents based on preset event detection thresholds.

Event platform:

The event platform is responsible for setting the detection parameters of microwave detectors, video detectors and traffic analyzers, and receiving and processing event information.

The microwave detector collects target information in the monitoring area, tracks the movement information of each target and the lane on which the target is traveling in real time, and sends the target information to the traffic analyzer; the video detector captures road video information in the monitoring area in real time, and The collected video information is sent to the traffic analyzer, and the target motion information includes position and speed.

The traffic analyzer receives the target information tracked by the microwave detector in real time, and determines the traffic event according to the preset event detection threshold; if the target is determined as a traffic event, the microwave detector sends a trigger signal to the video detector, and the video detector responds The target captures and sends the combined image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target's traffic incident and sends it to the event platform together with the target information and images detected by the detector, and the event platform finally pushes the traffic event alarm information; if The target is not judged as a traffic incident, and the traffic analyzer saves the road video information of the monitored area and does not report it to the incident platform.

When the microwave detector detects that a target enters the monitoring area, it sends a trigger signal to the video detector through the RS485 interface. The video detector receives the trigger signal, and the video detector will identify the feature information of the captured image and compare the image The recognition result is sent to the traffic analyzer; the image feature information includes license plate and vehicle type.

When the microwave detector detects that the target leaves the monitoring area, it sends a trigger signal to the video detector through the RS485 interface. The video detector receives the trigger signal, captures the target image and sends the image to the traffic analyzer.

The trigger signal includes: the location of the target and a time stamp; the video detector directly captures and recognizes the image of the target according to the target location; the time stamp is the time when the target enters the monitoring area, the time when it leaves the monitoring area, and the occurrence of violations time.

After the event platform receives the video information and image recognition results sent by the traffic analyzer, it displays information related to the traffic event.

As shown in Figure 2: A method for detecting multi-target traffic incidents in a large area includes the following steps:

The first step: system installation and adjustment and event detection threshold setting: firstly set up microwave detectors, video detectors, and traffic analyzers on the road members, and adjust the elevation, horizontal and tilt angles of the microwave detectors and video detectors , Until the area that needs to be monitored can be fully detected. The microwave detector and the video detector are connected by RS485 interface, the microwave detector and the traffic analyzer are connected by RJ45 interface through the switch, the video detector is connected by the RJ45 interface through the switch, and the traffic analyzer and the event platform realize data through the network. transmission.

The traffic analyzer has the NTP server function, which can ensure the complete synchronization of the traffic analyzer, microwave detector, and video detector.

Set the monitoring area and event detection threshold in the traffic analyzer, where the event detection parameters include lane, vehicle type, speed limit, etc.

Step 2: The traffic analyzer receives the data and completes the target complete information matching:

The microwave detector collects the target entering the monitoring area and sends the motion information and lane to the traffic analyzer, and sends a trigger signal to the video detector, where the motion information includes the position and speed, and the trigger signal includes the target's position and time stamp; video detection The device receives the trigger signal to capture the target image, and recognizes the captured image feature information, and at the same time sends the feature information image to the traffic analyzer, where the feature information includes the license plate and vehicle type; the traffic analyzer receives the information and video of the microwave detector The information of the detector is matched with the information of the same target, and the complete information of the target is generated. The complete information of the target is the position, speed, vehicle type and driving lane of the target.

The entire matching process of the target:

From the microwave detector coordinate system to the video detector coordinate system through rotation and translation, and finally project from the video detector coordinate system to the image coordinate system, the target in the microwave detector coordinate system and the target in the image coordinate system can be obtained. match. The ID (identifier), speed, location, vehicle type, and license plate number of the target are one-to-one correspondence, and finally all the information of the target is displayed on the image. Refer to Figure 3 for the specific matching display.

Step 3: The microwave detector and the video detector respectively monitor the traffic data and road images in real time, and send them to the traffic analyzer in real time

1) The microwave detector detects the motion information of the target vehicle on the road in real time, and tracks the motion information of all targets in the monitoring area in real time; this motion information includes position and speed until the target leaves the detection area, and sends the target motion information together with the time stamp to Traffic analyzer.

The microwave detector sets an identifier ID for each target vehicle that enters the monitoring area (the identifier ID is a custom number for the target vehicle). The microwave detector tracks the target by locating the coordinate position and speed of the vehicle in real time, and realizes the trajectory tracking detection of the vehicle in the monitoring area.

2) When a target enters the monitoring area, start the video detector to record the road image, the video detector captures the road image, obtain the characteristic information of all the targets in the monitoring area from the road image, and separate the target to be recognized from the background, according to The geometric shape, texture characteristics, etc. analyze the vehicle model (for example, car, bus, passenger car, etc.), identify the lane in which the vehicle travels, and identify the license plate number. And send the collected image information, the identified characteristic information (including the model, the driving lane and the license plate number) together with the time stamp to the traffic analyzer;

3) When the microwave detector detects that a target enters or leaves the monitoring area, it sends a trigger signal to the video detector through the RS485 interface, and the video detector receives the signal and captures the target.

Step 4: The traffic analyzer determines the event of the target tracked by the microwave detector:

The traffic analyzer receives the target information of the microwave detector in real time to determine traffic incidents according to the preset event detection threshold. The determination method is that the traffic analyzer compares the lane, vehicle type and driving speed of each target with the set event detection threshold. Determine whether a traffic incident occurs. When the traffic incident is not established, the traffic analyzer saves the road video information of the monitored area and does not report to the event platform; when the traffic incident is established, the microwave detector sends a trigger signal to the video detector, and the video detector captures the target , And send the captured image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target's traffic incident and sends it to the event platform together with the target information and images detected by the detector.

Step 5: After the event platform receives the video and pictures from the traffic analyzer, it displays the event alarm and event-related information for further processing.

In the traffic law, the speed limit for each type of vehicle is also different. For example, the high-speed limit is 120km/h for cars and 100km/h for buses. The purpose of vehicle type recognition by traffic analyzer is to detect the target speed and combine with it. The vehicle model determines whether it is speeding. And according to the planning of each road, the speed limit value is different, this system can be adjusted according to the actual situation of the speed limit value. For example, the speed limit for small cars is 120 and the speed limit for large cars is 100 on high speeds. Some speed limits on urban roads are 70 and some are 60. The speed limits set on different roads are different.

In the traffic law, each lane on a high-speed requires different vehicle types, for example, large trucks drive on the outer lane.

After the event detection threshold is set in the first step, within the detection range, it is calculated in real time whether it exceeds the maximum speed limit of the vehicle type. If the speed limit is not exceeded, the speeding event is not established and does not need to be dealt with; if the maximum speed limit of the vehicle is exceeded, the speeding event is established for further processing.

After the speeding event is established, the traffic analyzer tracks the speeding target in real time, and changes the color of the label box of the target that has the speeding event until the target stops speeding or leaves the monitoring area. Then generate three time records, the first is the time t ₁ _{when the target enters the detection area, the second is the time t 2} when the target is detected for the first time overspeed, and the third is the time t _{3 when the} target leaves the monitoring area; Then intercept the video of the time period from _{t 2} to t ₃ _{and the pictures of the three time nodes t 1} , t ₂ , and t ₃ ; finally, send the video and 3 pictures of the event to the event platform, and push the speeding alarm information at the same time.

After the road occupation event is established, track the vehicle target in real time until the target changes the correct lane or leaves the monitoring area, and then generates three time records, the first is the time t ₁ when the target enters the detection area, and the second is the first detection lane target to the traveling time t _2, a departure time of the third monitoring target region t _3; t ₂ is then taken to the time t ₃ videos and t _1, t _2, t ₃ time three nodes The video and 3 pictures of the incident were finally reported.

Step 6: Event alarm and event information display:

After the event platform receives the video and pictures from the traffic analyzer, it displays the event alarm and event-related information for further processing.

The present invention changes the existing detection method of taking photos and obtaining evidence at a certain fixed detection line. By adopting a microwave detector to detect the speed of multiple targets in multiple lanes in a large area in real time and a video detector to capture evidence, the detection area can be realized. One-to-one correspondence between internal detection targets and event records, realizing the detection of speeding events at any position in multiple lanes in a large area, effectively avoiding missed or multiple inspections;

The invention is composed of a single microwave detector, video detector, and traffic analyzer, which is convenient for debugging and maintenance;

The invention adds the determination of the vehicle type, regulates the vehicle type and speed of each lane on the expressway, and facilitates vehicle management and control.

The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the technical principles of the present invention, several improvements and modifications can be made. These improvements and modifications It should also be regarded as the protection scope of the present invention.

Claims

A large-area multi-target traffic incident detection system, which is characterized in that it includes a microwave detector, a video detector, a traffic analyzer, a switch, and an event platform;

RS485 serial communication mode is used for data transmission between the microwave detector and the video detector;

The microwave detector and the traffic analyzer, the video detector and the traffic analyzer are all interconnected through the switch, and the RJ45 network communication method is used for two-way data transmission;

The traffic analyzer and event platform use network communication for data transmission;

The microwave detector collects target information in the monitoring area and sends the target information to the traffic analyzer; the video detector captures road video information in the monitored area in real time, and sends the collected video information to the traffic analyzer; and When the target enters the monitoring area, the microwave detector sends a trigger signal to the video detector. The video detector receives the trigger signal to capture the image of the target, recognizes the captured image feature information, and sends the image with the feature information to the traffic at the same time Analyzer; the traffic analyzer receives target information, video information, and images with characteristic information, performs information matching of the same target, and generates complete target information;

The traffic analyzer receives the target information of the microwave detector in real time, and determines the traffic incident according to the preset event detection threshold; if the target is determined as a traffic incident, the microwave detector sends a trigger signal to the video detector, and the video detector detects the target Take a snapshot and send the combined image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target's traffic incident and sends it to the event platform along with the complete target information and images detected by the detector, and the event platform finally pushes the traffic event alarm information; if If the target is not judged as a traffic incident, the traffic analyzer saves the road video information of the monitored area and does not report the incident platform;

After the event platform receives the video information, target information, and images sent by the traffic analyzer, it displays information about the traffic event.
The large-area multi-target traffic incident detection system according to claim 1, wherein the target information includes target motion information and target driving lane, and the target motion information includes the speed and position of the target, and The characteristic information of includes the license plate and model.
The large-area multi-target traffic incident detection system according to claim 1, wherein when the microwave detector detects that a target enters the monitoring area, it sends a trigger signal to the video detector through the RS485 interface. Upon receiving the trigger signal, the video detector will identify the feature information of the captured image and send the result of the image recognition to the traffic analyzer.
The large-area multi-target traffic incident detection system according to claim 1, wherein when the microwave detector detects that the target leaves the monitoring area, it sends a trigger signal to the video detector through the RS485 interface, and the video detector receives When the trigger signal is reached, the image of the target is captured and sent to the traffic analyzer.
The large-area multi-target traffic incident detection system according to claim 1, wherein the trigger signal includes: the location of the target and a time stamp; the video detector directly captures the image of the target according to the target location and recognizes ; The timestamp is the time when the target enters the monitoring area, the time when it leaves the monitoring area, and the time when the traffic incident occurs.
A method for detecting multi-target traffic incidents in a large area is characterized in that it comprises the following steps:

S1: Deploy microwave detectors and video detectors on the rods, adjust the time of the two through a traffic analyzer, and set the event detection threshold;

S2: The microwave detector tracks the target information on the road in real time and sends the target information to the traffic analyzer; the video detector records the road video information of the monitored area in real time, and the video detector receives the trigger signal of the microwave detector and Identify the image information and send the identified image information to the traffic analyzer;

S3: The traffic analyzer receives the target information tracked by the microwave detector in real time, and determines the traffic incident based on the preset event detection threshold. The traffic analyzer compares the complete information of each target with the set event detection threshold to determine whether traffic occurs When the traffic incident is not established, go to S4; when the traffic incident is established, go to S5;

S4: The traffic analyzer saves the road video information of the monitored area, and does not report to the event platform;

S5: The microwave detector sends a trigger signal to the video detector, the video detector captures the target, and sends the captured image to the traffic analyzer; the traffic analyzer intercepts the video segment of the target where the traffic incident occurred and detects it together with the detector Send the complete target information and images to the event platform;

S6: The event platform pushes traffic event alarm information.
The method for detecting a large-area multi-target traffic incident according to claim 6, wherein the complete information includes target information and characteristic information, the target information includes target motion information and target driving lanes, and the The target motion information includes the speed and position of the target, and the feature information includes the license plate and car model; when the microwave detector detects that a target enters the monitoring area, the microwave detector sends a trigger signal to the video detector, and the video detector captures the target , And perform feature information recognition on the image, and send the image recognition result to the traffic analyzer;

When the microwave detector detects that the target leaves the monitoring area, the microwave detector sends a trigger signal to the video detector, and the video detector receives the trigger signal, captures an image of the target and sends the image to the traffic analyzer.
The method for detecting multi-target traffic incidents in a large area according to claim 6, wherein the traffic incidents include speeding incidents and road-occupying driving incidents.
The method for detecting traffic incidents with multiple targets in a large area according to claim 6, characterized in that, in S5, when it is judged that a traffic incident occurs, the target is tracked in real time until the target stops the traffic incident or leaves the monitoring area, and then generates three Time record, the first is the time t 1 when the target enters the detection area, the second is the time t 2 when the target is first detected a traffic incident, and the third is the time t 3 when the target leaves the detection area; then intercept The video of the time period from t 2 to t 3 and the pictures of the three time nodes of t 1 , t 2 , and t 3 ; finally, the video and 3 pictures of the traffic incident are reported.