System Analysis of Vehicle Violation Detection and Evidence Collection Using High Definition Video

The biggest difference between the off-site traffic enforcement model and the traditional law enforcement model is that once a violation occurs, visual data can be recorded. The perpetrator does not immediately know that he has been recorded at the scene of the violation, but only accepts investigation and punishment according to procedures after the visual data is reviewed and becomes administrative penalty evidence. In addition, from the perspective of the characteristics of evidence, the outstanding feature of visual data is that it can vividly and intuitively reflect the facts of the case, which is irreplaceable and incomparable to other evidence. The off-site law enforcement model of road traffic management refers to a new law enforcement model that uses video recording equipment and other means to collect visual data, record road traffic violations and violations, and punish them according to law.

The vehicle violation evidence collection system (abbreviated as: violation evidence collection system) is widely used at road intersections. It can effectively prevent traffic violations/violations such as running red lights, driving in the wrong direction, crossing lanes, changing lanes, etc., and reduce the traffic accidents caused by them.

The development process of the violation evidence collection system

Traditional traffic violation evidence collection system

The traditional system uses a ring-shaped coil vehicle detector to detect the driving status of the vehicle. It is composed of a coil vehicle detector, a signal light state detector, a camera, a centralized controller, a fill light, an intersection control host, etc. It captures illegal vehicles in the monitoring area, continuously captures multiple pictures and performs license plate recognition to form complete forensic data information.

The coil-type vehicle detector collects vehicle information, and the signal light state detector collects the signal light state. The outputs of these two detectors are connected to the centralized controller. The centralized controller determines whether the vehicle is in a red light violation state based on the vehicle position information and the signal light state. If it is in a violation state, a trigger pulse is sent to the camera, and the camera captures the image and transmits the image to the intersection control host. The intersection control host processes the image and performs license plate recognition, image synthesis, information superposition and other processing. After the processing is completed, the result is transmitted to the monitoring center through the transmission link, as shown in Figures 1 and 2.

▲Schematic diagram of the evidence collection system

The system has the characteristics of simple structure and high vehicle detection rate, but it can only collect evidence for violations such as running red lights, and cannot detect and collect evidence for violations such as driving in the wrong direction, crossing the line, changing lanes, etc. At the same time, due to the problems of easy damage to the annular coil, difficulty in repair, and damaged road surface caused by buried coils, its usage is gradually decreasing.

Video detection and evidence collection system

The system uses video detection to detect motor vehicles in real time. When a motor vehicle violates traffic rules such as running a red light, driving against traffic, or crossing a line, the camera captures multiple pictures and sends them to the control host. The controller host processes the pictures and uploads them to the control center. The computer system in the control center receives the data from the field equipment and stores it in the database after processing.

Standard definition video detection and evidence collection system

Since standard definition cameras have a small number of pixels (usually 400,000 pixels), vehicle license plates cannot be recognized when the image covers the entire intersection. For this reason, lane close-up cameras are needed to capture vehicle license plates, and panoramic cameras are used to capture images of vehicle violations.

The standard definition violation evidence collection system realizes the video detection of vehicles and can effectively collect evidence for violations such as driving against traffic, crossing the line, changing lanes, etc. However, due to the low pixel count of standard definition cameras, multiple cameras must be used to work separately, which leads to problems such as complex system structure and easy defects in violation evidence.

High-definition video detection and evidence collection system

With the rapid development of electronic technology , camera technology has entered the era of high-definition. At present, the number of pixels of high-definition cameras has reached more than 2 million. The use of high-definition cameras can achieve one high-definition camera covering one traffic direction, realizing the integration of close-up cameras and panoramic cameras. In a single picture, not only the panoramic view of the intersection and the process of vehicle violations can be displayed, but also the vehicle license plate can be clearly identified, solving the problem of defects in the evidence of violations by standard-definition cameras. [page]

Functions and trends of high-definition systems

High-definition full video detection and evidence collection system

System Features

The system uses high-definition video detection to detect motor vehicles and signal light colors in real time. When a motor vehicle violates traffic rules such as running a red light, driving in the wrong direction, or crossing a line, the high-definition camera captures multiple pictures to the controller, which processes the pictures and uploads them to the control center. The computer system in the control center receives the data from the field equipment, processes the data, and stores it in the database, as shown in Figures 3 and 4. Since video detection is implemented for motor vehicles and signal light colors, the system structure is simple, reliable, and easy to maintain.

▲Front-end layout diagram

System structure and function

The system consists of front-end acquisition/data processing equipment, transmission links, control center, etc.

The system can capture passing vehicles and illegal/violation vehicles in the monitoring area, continuously capture multiple process pictures, and automatically identify vehicle license plates to form complete forensic data information. Video detection can also be used to detect the color of signal lights, which can adapt to multi-phase signal lights and does not need to collect signal light color signals through electrical circuits. It can automatically determine the ambient brightness and automatically adjust the exposure parameters of the camera to keep the image in the best state and achieve the best recognition and monitoring effect. It can monitor the traffic status of vehicles at intersections. When there are violations such as running red lights, driving in the wrong direction, crossing the line, changing lanes, etc., it can continuously capture multiple process pictures, process the pictures through the controller, and upload the picture data to the control center.

System Advantages

The HD system has a simple structure. Only one HD camera is required for one traffic direction, without the need to connect other signal equipment.

The video detector can accurately track the movement trajectory of the vehicle, determine whether the vehicle is going straight, going in the opposite direction, turning left or right, record the vehicle's front and rear positions at the lane stop line, the direction of the vehicle crossing the lane divider, etc., without the need for external vehicle detectors (ring coils, geomagnetic rods, microwave, ultrasonic and other vehicle detectors).

The system can detect the color status of traffic lights through video, without the need for external red light detectors or cables to connect to traffic lights.

The controller adopts high-speed multi-channel DSP with powerful processing capability and is equipped with embedded operating system, dedicated device driver, efficient TCP/IP protocol stack, fuzzy recognition algorithm, etc. It can complete multiple functions such as video image analysis, target discrimination, speed detection, and license plate recognition.

The controller has extremely high processing power, which enables real-time processing of massive data. The application of advanced computer vision algorithms can process and identify images frame by frame during vehicle driving, effectively improving the equipment's ability to adapt to complex environments. The system can capture multiple valid frames, detect/identify each frame, and give the detection/identification results through the internal evaluation mechanism.

The upper left corner of the image can be overlaid with the following information: vehicle inspection time, number, vehicle type, lane number, driving type, whether it is a violation, whether it is driving in the wrong direction, the status of the vehicle's signal light before the stop line, the status of the vehicle's signal light after the stop line, the name of the intersection, the direction of the intersection, etc., as the basis for the penalty for violation. [page]

System application advantages

The system comes with a large-capacity hard disk that can store 800,000 checkpoint vehicle data or 200,000 violation penalty pictures. Depending on the traffic volume at the intersection, it can provide 15-45 days of historical data storage. When the transmission link with the monitoring center is interrupted, the controller can perform normal identification and storage; when the data link is restored, the stored historical data of any time period can be downloaded.

The controller can record high-definition video, and the frame rate can be set, the default is 2~4 frames per second.

When vehicles are allowed to pass, all vehicles passing normally are captured, recorded, and license plate recognized to realize the checkpoint function. The data records of the checkpoint include: 2 vehicle images, 1 small color image of the license plate, 1 license plate binary image, and 1 license plate recognition result.

According to the vehicle's movement trajectory and the color of the signal light, it is automatically determined whether the vehicle is in violation of traffic rules, and evidence is collected from the illegal vehicle. According to the standard, the evidence collection content of the illegal vehicle includes:

▲Forensic system architecture diagram

• The image can clearly identify the time of violation, red light signal status, license plate number, vehicle type, and lane stop line;

• Picture coding complies with the requirements of ISO/IEC 15444-2000;

• The records are finally synthesized into one picture: including three small pictures and one large picture, with superimposed time, location, direction, lane information, etc., as law enforcement evidence. The picture has a watermark function to prevent tampering;

• It can adapt to complex climate and lighting conditions, and can still maintain a high license plate recognition rate on cloudy days, rainy days, and strong direct or backlight. The license plate recognition rate is particularly high at night.

In addition, since the system uses the TCP/IP protocol to access the external network system, equipment upgrades, maintenance, and parameter adjustments can be completed online on any computer terminal in the entire network, greatly improving maintenance efficiency and reducing maintenance costs.