Please use this identifier to cite or link to this item:
Title: Tracking of multiple objects using the PHD filter
Keywords: Object tracking, source localization, random set, Bayes filter, PHD filter, GMPHD filter
Issue Date: 30-Sep-2008
Citation: PHAM NAM TRUNG (2008-09-30). Tracking of multiple objects using the PHD filter. ScholarBank@NUS Repository.
Abstract: Multiple object tracking is an important part of many applications. However, it is challenging due to the data association between objects and measurements. Based on the random finite set, the probability hypothesis density (PHD) filter operates on the single-object state space and avoids the data association. In the first part of the thesis, we propose a method to maintain track continuity in the Gaussian mixture PHD (GMPHD) filter. The identifications of objects are maintained during the tracking period with our method. In the second part, we employ the GMPHD filter in the multiple-speaker tracking. The method is efficient for real-time tracking of multiple speakers in a reverberant room. In the third part, we propose a PHD recursion for multiple-object tracking with color measurements. Lastly, we extend the method in the third part to multiple-camera multiple-object tracking. The results show that the proposed method is better than the stereo epipolar matching.
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
PhamNT.pdf2.53 MBAdobe PDF



Page view(s)

checked on Dec 16, 2018


checked on Dec 16, 2018

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.