Viswanathan, V
(2015)
Object Localization and Tracking in 3D.
Masters thesis, Indian Institute of Technology Hyderabad.
Abstract
The field of Computer Vision has repeatedly been recognized as an intellectual frontier whose boundaries of applicability are yet to be stipulated. The work attempts to demonstrate that vision can achieve an automatic localization and tracking of targets in a 3D space. Localization of targets has gained importance in the recent past due to the myriad of applications it plays a significant role in. It is analogous to detection of objects in a video sequence in the image processing domain. This work aims to localize a target based on range measurements obtained using a network of sensors scattered in the 3D continuum. To this end, the use of the biologically inspired particle swarm optimization(PSO) algorithm is motivated. In this context, a novel modification of PSO algorithm is proposed that leads to faster convergence, and eliminates the ip ambiguity encountered by coplanar sensors. The initial results over several simulation runs highlight the accuracy and speed
of the proposed approach.
Actions (login required)
 |
View Item |
Altmetric
Altmetric