Title :
Improvement of the tracking accuracy through the use of the support vector machine
Author :
Meetit, D. ; Kasetkasem, T.
Author_Institution :
Electr. Eng. Dept., Kasetsart Univ., Bangkok, Thailand
Abstract :
This paper extend the work by Aslam [7] by adding the direction information of the moving objection. The goal is to track the moving object by using binary sensors whose observations indicates whether or not the object is moving toward or moving away form. Aslam´s work mentioned the use of the moving direction of the object but did not give the specific detail on how to obtain the moving direction information. In this paper, we propose the technique to extract the moving direction of the observations of binary sensors. Here, the support vector machine(SVM) is employed to estimate the moving direction of the object since we observe that the optimal hyperplane resulting from the SVM is likely to be perpendicular to the moving direction. Because the same observation as proposed by Aslam is used in our approach, no additional communication between sensor nodes are needed. Our results show that, with this simple approach, the modest to high performance improvement can be achieved.
Keywords :
Clustering algorithms; Data mining; Global Positioning System; Object detection; Particle filters; Particle tracking; Predictive models; Support vector machines; Target tracking; Wireless sensor networks; binary sensor; particle filter; target tracking; wireless sensor network;
Conference_Titel :
Electrical Engineering/Electronics Computer Telecommunications and Information Technology (ECTI-CON), 2010 International Conference on
Conference_Location :
Chiang Mai, Thailand
Print_ISBN :
978-1-4244-5606-2
Electronic_ISBN :
978-1-4244-5607-9