Identification process
Automatic license plate recognition is a pattern recognition technology that uses dynamic video or static images of vehicles to automatically identify license plate numbers and license plate colors. Its hardware base generally includes trigger equipment (monitoring whether the vehicle enters the field of vision), camera equipment, lighting equipment, image acquisition equipment, processing machine (such as computer) for identifying license plate numbers, etc., and its software core includes license plate positioning algorithm, license plate character segmentation algorithm and optical character recognition algorithm. Some license plate recognition systems also have the ability to determine whether there is a car through video images, which is called video vehicle detection. A complete license plate recognition system should include several parts such as vehicle detection, image acquisition, and license plate recognition (as shown in Figure 1). When the vehicle detection part detects that the vehicle is arriving, the image acquisition unit is triggered, and the current video image is collected. The license plate recognition unit processes the image, locates the position of the license plate, and then divides the characters in the license plate for recognition, and then composes the license plate number output.
Vehicle Detection
Vehicle detection can be carried out in a variety of ways, such as buried coil detection, infrared detection, radar detection technology, and video detection. Video detection avoids damage to the road surface, eliminates the need for external inspection equipment, eliminates the need to correct the trigger position, saves money, and is more suitable for mobile, portable applications.
The system needs to have a high processing speed and an excellent algorithm to realize image acquisition and processing without losing frames. If the processing speed is slow, it leads to frame loss, so that the system cannot detect the vehicle with fast driving speed, and it is also difficult to ensure that the recognition processing starts at a position that is conducive to recognition, which affects the recognition rate of the system. Therefore, it is technically difficult to combine video vehicle detection with automatic license plate recognition.
Number identification
In order to perform license plate recognition, the following basic steps are required:
1) License plate positioning, positioning the location of the license plate in the picture;
2) License plate character segmentation, the characters in the license plate are separated;
3) License plate character recognition, the segmented characters are recognized, and finally the license plate number is formed.
In the process of license plate recognition, the recognition of license plate color is based on different algorithms, which may be realized in the above different steps, and usually cooperate with and verify each other with license plate recognition.
1. License plate positioning
In the natural environment, the background of the car image is complex and the lighting is uneven, and how to accurately determine the license plate area in the natural background is the key to the whole recognition process. Firstly, a large-scale correlation search is carried out on the collected video images, and several regions that conform to the characteristics of automobile license plates are found as candidate areas, and then these candidate areas are further analyzed and evaluated, and finally the best area is selected as the license plate area, and it is separated from the image.
2. License plate character segmentation
After the positioning of the license plate area is completed, the license plate area is divided into a single character and then recognized. Character segmentation generally uses the vertical projection method. Since the projection of the characters in the vertical direction must be near the local minimum value between the characters or in the gaps within the characters, and this position should meet the character writing format, characters, size restrictions and some other conditions of the license plate. The vertical projection method has a good effect on the character segmentation in the car image in the complex environment.
3. License plate character recognition
The main methods are based on template matching algorithm and artificial neural network algorithm. Based on the template matching algorithm, the segmented characters are first binary and their size is scaled to the size of the template in the character database, and then all the templates are matched, and the best match is selected as the result. There are two algorithms based on artificial neural networks: one is to extract the features of the characters first, and then use the obtained features to train the neural network allocator; Another method is to directly input the image into the network, and the network will automatically realize feature extraction until the recognition result is achieved.
In practical applications, the recognition rate of the license plate recognition system is also closely related to the quality of the license plate and the quality of the shooting. The quality of the license will be affected by various factors, such as rust, defacement, paint peeling, font fading, license plate obscured, license plate tilting, bright reflection, multi-license plate, fake license plate, etc.; The actual shooting process can also be affected by factors such as ambient brightness, shooting method, vehicle speed, and so on. These influencing factors reduce the recognition rate of license plate recognition to varying degrees, which is also the difficulty and challenge of the license plate recognition system. In order to improve the recognition rate, in addition to continuously improving the recognition algorithm, we should also find ways to overcome various lighting conditions to make the collected images the most conducive to recognition.
Schematic diagram of the license plate recognition system
The license plate recognition PTZ camera captures the license plate through light suppression shielding, electronic shutter adjustment, wide dynamic function, etc.:
Strong light shielding: On the basis of the low-light color camera, the brightest part of the image is blocked through the function of the software. In traffic monitoring, the strong light of the headlights can generally be blocked, so that the license plate can be captured more clearly. However, the biggest disadvantage of this camera is that the software is not clear, the brightest part of the image is not clearly defined, the license plate number may also be blocked, and it cannot handle the capture of high-speed moving objects. At present, domestic cameras have done more in terms of strong light shielding, and the effect is different from all parties.
ADJUSTABLE ELECTRONIC SHUTTER: FOR HIGH-SPEED MOVING OBJECTS (THE SPEED OF CARS ON THE HIGHWAY IS GENERALLY MORE THAN 70KM/H), YOU CAN REDUCE THE ELECTRONIC SHUTTER SPEED TO ACHIEVE CLEAR CAPTURE, WHICH IS GENERALLY ADJUSTED MANUALLY. However, the biggest problem with this method is that the illumination is different during the day and at night, and it is necessary to set two shutter speeds to adapt to the monitoring during the day and at night.
Wide dynamic function: This is the best way to solve the impact of car lights on capture, wide dynamic range was first proposed by Panasonic. When the background light is too bright, ordinary cameras cannot solve the display problem of light and dark images.
As the most important means of determining the identity of vehicles in the field of intelligent transportation, license plate recognition technology has played a great role in realizing the intelligent management of traffic, and there is the penetration of license plate recognition technology in various work.
1. Monitoring and alarming
For the vehicles included in the "blacklist", such as: wanted or lost vehicles, non-payment vehicles, vehicles that have not been inspected, hit-and-run and illegal vehicles, etc., only need to enter their license plate numbers into the application system, license plate recognition equipment installed at designated intersections, bayonets or carried by law enforcement officers at any time as needed, the system will read all the license plate numbers of passing vehicles and compare with the "blacklist" in the system, once the designated vehicle is found to immediately send out alarm information. The system can work non-stop all day without fatigue and has a very low error rate; It can adapt to high-speed vehicles; It can complete the task during the exercise of the vehicle without affecting the normal traffic; During the whole monitoring process, the driver will not be aware and the confidentiality is high. The application of such a system would greatly improve the efficiency of law enforcement.
2. Penalties for speeding violations
License plate recognition technology combined with speed measurement equipment can be used for vehicle speeding violation penalties, generally used on highways. The specific application is: setting up speed monitoring points on the road, capturing speeding vehicles and identifying license plate numbers, and sending license plate numbers and pictures of illegal vehicles to each exit; Set up penalty points at each exit, use the license plate recognition device to identify the passing vehicle and compare the number with the number of the speeding vehicle that has been received, and once the number is the same, the warning device will be activated to notify the law enforcement officers to deal with it. Compared with the traditional speed monitoring method, this application can save police force, reduce the work intensity of law enforcement officers, and is safe, efficient and hidden, and the driver needs to remind himself not to speed at all times, which greatly reduces the accidents caused by speeding.
3. Vehicle access management
The license plate recognition equipment is installed at the entrance, the license plate number and entry and exit time of the vehicle are recorded, and the control equipment of the automatic door and barrier machine is combined to realize the automatic management of the vehicle. It can be applied to the parking lot to realize automatic hourly charging, and can also automatically calculate the number of available parking spaces and give prompts, so as to realize the automatic management of parking charges, save manpower and improve efficiency. When applied to the intelligent community, it can automatically determine whether the vehicle entering the community belongs to the community, and realize automatic timing and charging for non-internal vehicles. In some units, this application can also be combined with the vehicle dispatching system, automatically, objectively record the vehicle of the unit, H3 wisdom in the traffic management system using license plate recognition technology, to achieve non-stop, free card, effectively improve the efficiency of vehicle access.
Fourth, automatic release
The specified license plate information is entered into the system, and the system automatically reads the license plate of the passing vehicle and queries the internal database. For vehicles that need to be automatically released, the system drives the electronic gate or barrier to let it pass, and for other vehicles, the system will give a warning and be handled by the personnel on duty. It can be used for special units (such as military management areas, secret units, key protection units, etc.), road and bridge toll bayonets, high-end residential areas, etc.
5. Highway toll management
Install license plate recognition equipment at each entrance and exit of the expressway, identify the vehicle license plate when the vehicle enters, store the entrance data into the toll system, identify its license plate again when the vehicle arrives at the exit and call the entrance data according to the license plate information, combine the entrance and exit data to realize toll management. This application automates billing and prevents fraud and avoids the loss of receivables.
The highway has begun to implement the network toll, with the expansion of the scope of networking, the toll difference of different models is also getting higher and higher, the problem of driver evading the toll by changing the card halfway by using the loopholes of the existing toll system will become more and more prominent, and the use of license plate recognition technology is the fundamental method to solve such problems.
6. Calculate the travel time of the vehicle
In a traffic management system, the average travel time of a vehicle on a certain road can be used as a parameter to judge the congestion of the road. Install a license plate recognition device at the start and end points of the road, read all passing vehicles and transmit the license plate number back to the traffic command centre, based on which the command centre's management system can calculate the average travel time of the vehicle.
7. Automatic registration of licence number
The traffic supervision department has to deal with a large number of illegal vehicle pictures every day, and the license plate number is generally manually identified and then entered into the management system, which is a large workload and easy to fatigue and misjudge. The use of automatic identification can reduce the work intensity and greatly improve the processing speed and efficiency. This kind of function can be used for electronic police systems, road monitoring systems, etc.
The license plate recognition parking management system automatically recognizes and converts the image of the vehicle's license plate number taken by the camera at the entrance into a digital signal. The advantage of card and car correspondence is that the long-term rental card must be used in conjunction with the car, so as to eliminate the loopholes of the use of one card and multiple vehicles, and improve the efficiency of property management; At the same time, it automatically compares the incoming and outgoing vehicles to prevent theft. The upgraded camera system can capture clearer pictures and keep them as archives, which can provide strong evidence for some disputes. It is convenient for the management personnel to compare when the vehicle is out of the field, which greatly enhances the safety of the system.
8. License plate recognition has broad application prospects
With the continuous development of parking lot management system, license plate recognition technology has also begun to be applied to parking lots. However, according to the survey, because there are only a few domestic enterprises specializing in the development of license plate recognition technology for parking lots, so many parking lot license plate recognition technology is mostly copied from the early license plate recognition technology, and the two in the recognition environment, recognition angle, lighting environment, vehicle driving speed and other aspects of the big difference, but also led to the phenomenon of "not adapting" to the soil and water most of these license plate recognition systems. In this regard, many car parks have called for a "customized" license plate recognition system.
Statement: The article is my original, it is forbidden to reprint, if you have any questions, please contact me 13371621129 QQ: 377428678