Urban road monitoring of the electronic police system (1)

The rapid development in computer technology today, security has a long history of the industry has become a hot industry.

Although the value of the security industry also far and metallurgy, electric power, telecommunications, petroleum, chemical industry, machinery, etc. several big industry, but it IT, ITS other disciplines such as intelligent transportation technologies increasingly fusion, and to each of the 25% growth rate shows a better future.

As the urban road monitoring technology and ITS intelligent transportation technologies development and integration, "electronic police" has become a easing traffic stress, reduce traffic accidents and hazards reflected informationizaton an important initiative.

Electronic police enforcement penalties system with high technology content, you can "speeding, retrograde, red light running, stopped, pressing the yellow line, preemption bus lane, and a series of illegal phenomenon for accurate, stable, and automatic, 24 x 7 monitoring, enforcement and penalties. Electronic police presence can greatly alleviate violation leads to increased traffic accident and the police forces and police officers in the intensity of the conflict, the effective suppression of human traffic accident caused by the violation. Electronic police in some cities over the past few years, technology has been very mature, from early single red light to capture the number of illegal content and traffic flow, queue length, waiting time, lane share, the average speed of traffic information for statistical analysis of the monitoring system. Nowadays the main urban transport electronic police include: vehicle ran a red light violation the video monitoring system, vehicle speed monitoring system, retrograde stop vehicle monitoring system, bus lane monitoring system, non-motorized road traffic monitoring system, pressure double yellow lines monitoring system, emergency shutdown with traffic monitoring system, mobile vehicle inspection system, intelligent multimedia network license plate recognition systems, traffic flow monitoring system, etc.

Electronic police major core part is the vehicle detection technology, motor vehicle license plate recognition technology and video camera.

Vehicle video detection technology

Present in all major cities in the electronic police use of vehicle detector is coil inductor and video detector.

Buried inductive loop detector, because of its high reliability, high accuracy and low prices, in a few years ago has been widely used, but with the development of electronic technology, large-scale, VLSI development, powerful digital processing chip, vehicle video detection technology breakthrough and development of very mature. And coil detection technology superiority over and the money is beginning to be recognized in the industry, represents the future vehicle detection areas of development and application of direction.

Video vehicle detector is a video image analysis and computer vision technology for road vehicles running checking and analysis of the integrated system.

It can enter real-time analysis of traffic images, tracking of vehicles in the image, access to traffic data.

Video vehicle detector's main advantages:

1, the use of panoramic camera as video sensor, it does not require cutting, as well as the laying of pavement destroyed a long coil feeder cable, the greatest degree of shortening the time for road closures.

In addition, the location of the video virtual coil can be arbitrary, and inductive loop installed in a location that does not change depending on road conditions, otherwise duplicate any cutting pavement can seriously impact the life of the road.

2. video virtual coil position display can be adjusted according to the traffic, its coil features properties can also be set as needed, such as: exist, delay, delay, pulse and count.

Selection of different properties of coil can implement different control requirements, each channel video image can be set more than one such virtual coil.

3. Induction coil's life in about 2-5 years (depending on traffic and road surface temperature), and video detector life to more than 10 years.

4, the use of a digital signal processing (DSP) and advanced algorithms, video detection can face the pavement reflection, vehicle shadows, bad weather and the challenges of various lighting conditions, which are precise and reliable measurement data, the actual realization of inspection control, 24 hours a day.

Sketch map

Vehicle video detection technology in traffic data collection, statistics, analysis, application of the most complex, this kind of video traffic detector uses video processing and recognition technologies and construction of the Observatory, the ability to automatically extract several traffic flow parameters, automatic classification models, and can detect general traffic accidents and other special traffic accident.

The detector can be installed in the highway traffic roads or important traffic observation points, used in place of the traditional traffic Observatory. In addition, in the red light, speeding, forbid of mobile electronic police system, vehicle detection using video technology, and traditional underground Induction coil is being replaced.

Vehicle license plate recognition technology

Vehicle license plate recognition as a traffic monitor application of core technologies in multiple subsystems, such as the red light monitoring systems, monitoring systems, retrograde overspeed monitoring system, stop monitoring system, bus lane monitoring system, non-motorized road traffic monitoring system, pressure double yellow lines monitoring system, emergency shutdown with traffic monitoring system, mobile vehicle inspection system, and so on.

Intelligent multimedia network license plate recognition system widely used in automatic registration, passing vehicles, road tolls, vehicle security verification, community, parking management, etc.

System uses video real-time trigger mode detection capture, the ability to automatically detect, identify and verify driving or parking of vehicles by vehicle plate number.

You can now capture images and database information in a timely manner, when the discovery should stop the car, the system can be on the local machine and on timely alarm centre. The system uses the Advanced Fuzzy image processing technology, through program can very good implementation for overall tilt, vehicle license plate text italic, license plate of defaced and fuzzy, and so would the human eye is very difficult to distinguish between recognition of license plate numbers.

Vehicle license plate recognition process can be divided into license plate location, license plate pre-treatment, segmentation and character recognition four steps.

nter>

Figure 1 vehicle license plate recognition process

Implementation of functional and technical features

★ Accurate identification in different regions and different types of numbers.

★ Image automatically trigger mode without any external trigger mechanism.

¡Ï auto complete vehicle count, traffic statistics.

¡Ï on capture image has been able to keep up to date with the database information, when the discovery should stop the car, on your local machine and Center machine in a timely manner.

★ Built-in database management software to store, search and organize information on vehicle can automatically backup data and complete statistics reports.

★ In network environment to achieve synchronization of data around, you can monitor the health of the front-end system.

¡Ï on movement speed 180 km/h following car license plate for automatic recognition.

★ In good lighting conditions, license plate recognition rate is not less than 96%, in rainy days and night artificial illumination conditions, license plate recognition rate is not lower than 90%.

The system can identify the plate types include: ordinary civil automobile license plate, license plate for military vehicles (including armed plate), police car license plate

The system can identify vehicle type, draw a three-dimensional image of the vehicle.

* Capture images for less than 0.03 seconds, identify images less than 0.2 seconds.

★ The system adapts to all-weather conditions.

[1] [2]