Development Of Intelligent Video Surveillance Platform And Its Application And Implementation Of Vehicle License Plate Recognition System In Parking Lot

With the development of semiconductor chip, computer science and networkcommunication technology, the video surveillance is a key technique in securitysystem, which is developed from the first generation of full analog system to thesecond generation of semi-digital and fully digital systems, then to the thirdgeneration of integration and network systems. Nowadays, by means of the computervisual technology in artificial intelligence, the content in video monitor screen isprocessed by using computer to deal with digital image. Security video surveillance isdeveloping in the directions of the intelligent network video surveillance system.Based on the project of intelligent parking recognition system combined withSmart Visual Security in the Internet of Things Laboratory, the whole architecture oflarge-scale video surveillance platform in city level is designed. Combined with someconcrete illustrations of parking, the automatic vehicle recognition algorithm isdeveloped based on the thorough research to security video surveillance system.Video surveillance of the parking system is divided into the three-tier system ofmunicipal users, district-level users and parking lots users, which are usingon-demand access methods. On this basis, the paper put forward a parking IntelligentLicense Plate Recognition System with a two-terminal framework. It adopts the purevideo trigger to detect vehicle. Based on the virtual coil self-triggering, feature imagecontaining the license plate from the front is grabbed and saved, then to server byinternet for further processing. According to the principle of modular design, thegeneral-purpose embedded cameras with Hi3516as core processor is designed,including IPCamera wrapper design, interface board drawing, which the BootLoaderand Linux kernel and Porting operating of file system in the Hi3516developmentboard are realized, the environment of the software and hardware of the system arealso built, the cross compiler is installed, and the latest version of OpenCV2.4.2wastransplanted to the Hi3516embedded platform.At last, application of the whole system in the parking is realized throughprocessing of the image intelligent algorithm, and the performance requirements oflicense plate recognition system in the parking is satisfied by the field testing.