Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

Traffic video monitoring of certain key algorithm design and application

Author HeJing
Tutor WangGuoYou
School Huazhong University of Science and Technology
Course Pattern Recognition and Intelligent Systems
Keywords Video Surveillance Vehicle Detection Vehicle speed License Plate Recognition Marking detection
CLC TP391.41
Type Master's thesis
Year 2011
Downloads 67
Quotes 0
Download Dissertation

Traffic through the video surveillance system to capture video from a traffic surveillance target , target information extraction and analysis process reached on urban traffic monitoring and management. This paper focuses on intelligent traffic control system in the framework of a study on the system involved in key technologies such as image processing algorithm for vehicle detection , vehicle speed , and license plate recognition algorithm algorithm carried out relevant research , and their realization. In the system architecture , this paper presents a basic vehicle information element based system architecture . The system architecture can essentially reflects the basic objective of monitoring systems for the vehicle management objectives, the entire system from the vehicle license plate recognition process detects the vehicle unified information element attribute feature extraction management . Not only for the system provides better scalability, but also for system management provides a great convenience. In the vehicle detection algorithm research and implementation process , this paper mainly implemented based on background subtraction and frame difference motion region detection algorithm, and for road transport vehicles prone to adhesions , implemented based on concave point detection the adhesion segmentation method that can better deal vehicle detection frequently encountered shadows and adhesion problems . Vehicle detection algorithm in the study , we propose a road marking detection based vehicle detection algorithm, which is based on the world coordinate system to the image coordinate system conversion, omit the complicated camera calibration process, and according to the vehicle information element location attribute , you can calculate the image distance between any two points , thus greatly improving the accuracy of vehicle speed . In license plate recognition algorithm , there would be a vehicle license plate information element as a basic property , the paper realized based on morphological edge enhancement and license plate location algorithm based on Radon transform plate tilt correction algorithms. License plate character segmentation algorithm in the research process , we propose a template -based segmentation method barycenter . License plate character alphanumeric reference application characteristics to achieve a template matching based on multi -level alphanumeric classifier in the experiments and achieved good results.

Related Dissertations
More Dissertations