Title :
A Novel Strategy for Visual Tracking and Visual Information Extraction
Author :
Liu, Zhiyuan ; Lee, Bingbing ; Lee, Kai ; Jiang, Jing ; Lee, Meng ; Qiu, Jiejun
Author_Institution :
No. 49 Res. Inst., China Electron. Technol. Corp., Harbin, China
Abstract :
A novel strategy for the visual tracking and the information extracted problem is proposed which is for the case of maneuvering target. The strategy contains two methods. One is used to tracking, and the other one is used to extract the information of the vehicle. A non-linear estimation method using the particle filter to track objects is presented. During the tracking, a great deal of vehicle information is extracted at the same time, such as the type and velocity of vehicle. An improved Fisher linear discriminate is proposed to extract the license plate. By experiments, this strategy is proved much effectively, and the accuracy of tracking is improved greatly.
Keywords :
image motion analysis; information retrieval; object detection; particle filtering (numerical methods); principal component analysis; traffic engineering computing; Fisher linear discriminate; automatic traffic monitoring; motion detection; nonlinear estimation method; particle filter; principal component analysis; visual information extraction; visual tracking; Biomedical imaging; Data mining; Filtering; Licenses; Motion detection; Particle filters; Roads; Sensor fusion; Target tracking; Vehicles; Particle filtering; Tracking; Visual information processing;
Conference_Titel :
Measuring Technology and Mechatronics Automation, 2009. ICMTMA '09. International Conference on
Conference_Location :
Zhangjiajie, Hunan
Print_ISBN :
978-0-7695-3583-8
DOI :
10.1109/ICMTMA.2009.508