Factors Affecting Automatic License Plate Recognition and Problems Needing Attention During Purchase

Nowadays, license plate recognition technology has been widely used in urban traffic and parking lots, which has greatly improved the problems of traffic congestion and difficult parking. Of course, license plate recognition technology is not omnipotent, and some factors will affect its recognition effect. Below, Let's take a look at the factors affecting automatic license plate recognition and the problems that need to be paid attention to when buying. I. factors affecting automatic license plate recognition: 1. The influence of image on recognition. Only when the image is focused clearly, the recognition result can achieve a satisfactory effect.

When the size of the license plate in the image is 150x40 dot matrix, the characters are not easy to adhere, which is conducive to segmentation, and the character features after segmentation are obvious, which is conducive to recognition. 2. The impact of vehicle speed on images. The PAL video standard adopted in China is 25 images per second (the interval between images is 40 milliseconds). Suppose the depth of field (clear image range) of the camera lens is 1.0m. For the vehicle speed of 90 km / h (0.001 * 3600 / 0.04 = 90), the camera can only capture 1 frame of clear pictures. In order to get more clear images, we should try to increase the depth of field. The specific way is to increase the lens and reduce the size of CCD camera. 3. Impact of trigger device on image in Windows multitasking system, the task switching time is extremely unstable. According to the actual measurement, this time ranges from 0 to hundreds of milliseconds.

When the system CPU occupancy is high, the task switching time is very long. The CPU occupancy of video application system is relatively high. The trigger device actually uses the task query method to detect the state of the I / O port and capture the image when the set conditions are met. Because there is a certain delay in detecting the change of I / O port status, the captured image may not be the clearest. This situation is more obvious when the vehicle speed is high. 4. The influence of the vehicle model on the image is obvious whether the ground induction coil or infrared trigger is used. The trigger position of the crane and trolley may exceed 0.5m even at low speed. When it exceeds 1m, the captured image may not be the clearest. 5. Why is the recognition speed fast? On urban roads and national highways, the speed is generally 0 to 80 km / h. The actual vehicle arrest application system should reach this speed level, rather than limit the speed.

At the speed of 80km / h, the camera has only one clear image. The only way to accurately capture this frame of image is to capture and recognize the image frame by frame, which requires that the capture and recognition time shall not be more than 40ms. In order to ensure the normal operation of windows system, this time has to be reduced to less than 20ms. II. Points for attention when purchasing license plate recognition system: 1. Stability with the improvement of product algorithm and technology of operators, tigerwong believes that integration has long been an important topic, and the most important thing to pay attention to at this stage is stability. The definition of stability is: within a given speed range, the accuracy that must be achieved will not produce excessive error due to the influence of external environment. 2. The measured recognition rate almost everyone claims to have a high recognition rate.

However, in order to avoid shifting the responsibility for poor operation due to differences in product cognition between the two sides, users may wish to require field testing when purchasing the license plate recognition system, and the testing time should be more than two weeks, which can better judge whether the recognition results are exaggerated. Because Taiwan is a changeable environment, we should be able to grasp about 80% of the situations that may affect the recognition rate in the field in two weeks. If we only measure it for a day or even a few hours, we can't understand it. 3. Software and hardware architecture in addition, since license plate recognition is a system, the quality of software and hardware architecture will certainly affect the results presented. As for what kind of software and hardware is suitable for what kind of environment, it must vary according to the environment, because different application environments may not have the same requirements for recognition rate, which must rely on experience accumulation. With the wide application of license plate recognition, it has gradually become an important part of urban intelligent transportation and car parking management system, and has also prompted many parking lots to introduce automatic license plate recognition technology.

Although there are many parking lots equipped with license plate recognition technology in China, there are still some areas that are not installed and do not know much about it. Therefore, When purchasing the license plate recognition system, you need to know some basic knowledge. Don't buy it that doesn't work, spend unjustified money and delay the project. Tigerwong parking equipment supplier has focused on parking equipment for many years! If you have any questions about the parking system, welcome to consult and communicate.