What is AI intelligent license plate recognition technology?
AI intelligent license plate recognition technology is the application of computer video image recognition technology in license plate recognition. License plate recognition technology requires to obtain and recognize the target license plate number from the complex dynamic background. Through the acquisition of license plate number, image pre-processing, target feature extraction, character recognition of the license plate number and other technologies, the target license plate number, color characteristics and other information can be obtained.
In order to carry out license plate recognition, the following basic steps are needed:
1) Locate the license plate, locate the position of the license plate in the image; 2) License plate character segmentation, license plate character segmentation; 3) license plate character recognition, recognition of segmentation of the character, the final composition of the license plate number.
Due to the continuous development and progress of artificial intelligence technology, license plate recognition technology has gradually been applied to daily life scenes, such as intelligent parking lot, automatic billing, community vehicle access management, parking and violation monitoring, mobile criminal investigation. Moreover, AI license plate recognition technology is not only a key means of vehicle recognition combined with DSRC technology in unmanned toll collection system, but also plays an important role in highway vehicle management.
China’s standard car license plate is composed of Chinese characters, English letters and Arabic numerals. Current AI technology can recognize 99 percent of Chinese characters and 99.7 percent of letters and numbers. License plate recognition function based on computer vision and other technologies, the vehicle image or video image taken by the camera is analyzed, and the unique license plate number of each car is obtained, so as to complete the recognition process. Then match it with the vehicle and owner information in the database to realize the application in the real scene.
Application example of license plate recognition technology in parking scenario
The intelligent parking lot system integrates sensing smart card technology, computer network, video monitoring, image recognition processing and automatic control technology to automatically manage the vehicles in the parking lot. For example, including vehicle identity judgment, access control, automatic license plate recognition, parking lot retrieval, parking lot guidance, vehicle reminder, image display, model calibration, time calculation, charge query, voice dialogue, automatic take (receive) card and a series of scientific and effective operations.
The parking management system based on license plate recognition technology is combined with electronic unmanned charging system (ETC) to identify vehicles, which can realize automatic identification of vehicles, automatic charging, no vehicles through the intersection parking and other functions. In the management of parking lot, in order to improve the efficiency of vehicle import and export, automatic license plate recognition of vehicles does not need to levy parking fees (such as monthly trucks, internal vehicles free passage), can carry out the construction of unmanned express channel, change the traditional parking lot access management mode.
Vehicle detection can use embedded coil detection, infrared detection, radar detection technology, video detection and other methods. Using video detection can avoid road damage, do not need to add external detection equipment, do not need to modify the trigger location, save transformation and construction costs, suitable for mobile portable applications.
Application of video technology
License plate recognition based on video image processing technology, through installed in the road or gate bayonet cameras and image acquisition equipment will be real-time video information, through the real-time processing and analysis of video images to get all kinds of information, such as license plate color, license plate number, model, vehicle color, etc..
Because the license plate recognition parking system needs to detect the video target vehicle, it needs to have high processing speed and excellent AI algorithm, almost can not lose the frame to realize the image acquisition and processing. If the processing speed is slow, frame loss will occur, leading to the system can not detect the fast vehicle, difficult to ensure the license plate recognition processing, thus affecting the recognition rate of the system.
Compared with other traffic flow detection technologies, video-based vehicle recognition technology has the following advantages: 1) Video detection can detect a large area of traffic scenes; 2) Video sensors and other equipment, such as cameras, are easy to install and debug, and will not damage road facilities; 3) More traffic flow parameters can be collected by using video detection technology.
The License plate recognition function of the EasyCVR video fusion platform developed by TSINGSEE Qingxi Video is based on license plate location, license plate character segmentation and OCR recognition algorithm to quickly extract license plates from images and output the results. The recognition rate can also be guaranteed in complex environments such as rain and fog, complex light and night. Compared with the traditional card admission, license plate recognition is free of parking and window opening, which improves the user experience, reduces the cost of IC card loss and consumption in the parking lot, and improves the efficiency of car owners.