License plate recognition parking lot management system automatically recognizes the vehicle license plate image taken by the camera at the entrance and converts it into a digital signal. To realize one card and one car, the advantage of license plate recognition is that it can correspond the card with the car, so that management can be improved to a higher level. The corresponding advantage of the card with the car is that the long-term rental card must be used with the car, eliminating the loopholes in the use of one card with more cars and improving the efficiency of property management. At the same time, it automatically compares the vehicles entering and leaving to prevent theft. The upgraded camera system can collect clearer pictures and save them as files, which can provide strong evidence for some disputes. It is convenient for managers to compare vehicles when they leave the scene, which greatly enhances the security of the system.
1. Vehicle inspection
Vehicle detection can adopt many ways, such as buried coil detection, infrared detection, radar detection, video detection and so on. Using video detection can avoid damaging the road surface, without additional external detection equipment and correcting the trigger position, which saves money and is more suitable for mobile and portable applications.
The license plate recognition system with video vehicle detection function firstly collects a frame (field) signal in the video signal and digitizes it to get the corresponding digital image; Then analyze and judge whether there is a vehicle inside; If there is traffic, proceed to the next step of license plate recognition; Otherwise, continue to collect video signals for processing.
For video vehicle detection, the system needs high processing speed and excellent algorithm to realize image acquisition and processing without losing frames. If the processing speed is slow, it will lead to frame loss, which will make the system unable to correctly detect fast-moving vehicles. At the same time, it is difficult to ensure that the recognition processing will start at a position conducive to recognition, which will affect the recognition rate of the system. Therefore, it is technically difficult to combine video vehicle detection with automatic license plate recognition.
2. License plate number and color recognition
In order to perform license plate recognition, the following basic steps are required:
? Locating the license plate, namely locating the position of the license plate in the picture;
? License plate character segmentation, which separates the characters in the license plate;
? License plate character recognition, which recognizes the segmented characters and finally forms the license plate number.
In the process of license plate recognition, license plate color recognition is based on different algorithms, which may be realized in the above different steps, and usually cooperate with license plate recognition and verify each other.
(1) license plate location
In the natural environment, the background of automobile images is complex and the illumination is uneven. How to accurately determine the license plate area in the natural background is the key to the whole recognition process. Firstly, the collected video images are searched in a wide range, and several areas that meet the characteristics of automobile license plates are found as candidate areas. Then, these candidate regions are further analyzed and judged, and finally the best region is selected as the license plate region and separated from the image.
(2) License plate character segmentation
After the location of the license plate region is completed, the license plate region is divided into single characters and then recognized. Character segmentation generally adopts vertical projection method. Because the projection of characters in the vertical direction is bound to approach the local minimum at the gap between characters or inside characters, and this position must meet some conditions such as the character writing format, characters and size restrictions of the license plate. Vertical projection method has a good effect on character segmentation in automobile images in complex environment.
(3) License plate character recognition
At present, the main methods of character recognition are template matching algorithm and artificial neural network algorithm. Template-based matching algorithm first binarizes the segmented characters, scales their size to the size of templates in the character database, then matches all templates, and finally selects the best match as the result. There are two algorithms based on artificial neural network: one is to extract the features of the characters to be recognized, and then train the neural network distributor with the obtained features; Another method is to input the image to be processed directly into the network, and the network automatically realizes feature extraction until the result is recognized.
In practical application, the recognition rate of license plate recognition system is closely related to the quality of license plate and shooting quality. The quality of license plate will be affected by various factors, such as rust, stain, paint loss, font fading, license plate being blocked, license plate being tilted, highlighting and reflecting, multiple license plates, fake license plates and so on. The actual shooting process will also be affected by environmental brightness, shooting brightness, speed and other factors. These factors reduce the recognition rate of license plate recognition to varying degrees, which is also the difficulty and challenge of license plate recognition system. In order to improve the recognition rate, in addition to constantly improving the recognition algorithm, we should also find ways to overcome various lighting conditions to make the collected images most conducive to recognition.