What are the problems with license plate recognition system?
Decoiler Machine,Coil Decoiler,Steel Coil Decoiler,Coiler Decoiler Shandong EN FIN CNC Machinery Co., Ltd , https://www.shuofangcn.com
Although various camera manufacturers have developed corresponding products for road monitoring, they have made special improvements to the special requirements of traffic. However, due to environmental factors, the cameras are outdoors for a long time due to road monitoring. Therefore, they require not only high definition but also strong adaptability. Shenzhou Digital Ye Shuguang, for example, said that the use of license plate license plate recognition in the use of the environment is relatively more complex, some national roads, some provincial roads, some county roads, and different use of the vehicle's driving speed is different, which set the shutter speed of the camera The requirement is higher, not only to facilitate the user to install and debug, but also to set the shutter speed to be more accurate to adapt to different use places. For this application, the digital camera designed by China Digital AMPON has a Traffic traffic-specific function that can determine the shutter speed through a fixed mode for different road use scenarios, simplifying the user's debugging work. The shutter speed subdivision can also be applied to more occasions.
On the other hand, it is worth noting that the snow, snow, and foggy weather in the north have occurred in the winter. In this case, the visibility of the road surveillance camera will be very problematic. Therefore, many manufacturers have designed a fog-permeable function. However, China Digital Ye Shuguang said that there are many black and white fogging features on the market, but this will lose the color of the license plate. Therefore, ShenZhen Digital AMPON has specially developed color penetrating fog, through a unique detection circuit, it automatically detects the grayscale fog image data in the screen, performs unique processing for the fog screen, adjusts the dynamic range curve of the graphic in real time, and strengthens the color reproduction. Degree, providing a more realistic monitoring picture.
High clarity but brings other problems
The current intelligent transportation system has gradually transitioned from standard-definition systems to high-definition systems. The advantages of high-definition systems are self-evident, but Tian Di Weiye Sun Tao believes that high-resolution pictures may have multiple license plate recognitions in the picture due to the wide coverage of the pictures. problem. But what is more important is that if the amount of high-definition data is too large, not only will the processing speed be too slow, but it will also be difficult to realize high-definition video stream recognition due to excessive resource requirements. These are license plate recognition systems that face high-definition systems. New issues.
Damaged old license plate recognition is more difficult
After a license plate is used for a few years, it will inevitably cause pollution, wear, etc., and it is difficult to guarantee that vehicles traveling on the road are all standard and clean license plates. Therefore, in the actual environment, how to improve the face of damaged old license plates? The recognition ability of the license plate recognition system is also a problem that needs to be solved.
Han Hui Jiang Hui, for example, said that in Shanxi, Shaanxi and other long-term coal pull vehicles, due to dirty, will block a part of the license plate, or intentionally block the license plate. In view of this situation, the license plate recognition system is unrecognizable, because it is only an identification technology for existing images and does not include perspective capability. Therefore, there is a case where there is both interference and defilement in the video stream. To locate and identify license plates, the current license plate recognition system needs further improvement and development.
Beijing Wenan He Maolin stated that the current solution to this problem is to improve the recognition rate of the license plate recognition system. Currently, the main method adopted is to judge the vehicle body first in the video stream, and then to initially locate the license plate and After making a judgment, after the license plate is accurately positioned, the license plate is subjected to character and number recognition through an algorithm to obtain information on the license plate. However, this method mainly depends on the accuracy of the algorithm, so the quality of the algorithm is crucial.