Multiple object tracking using a transform space

M. Li*, J. Li, A. Tamayo, L. Nan

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

9 Downloads (Pure)

Abstract

This paper presents a method for multiple object tracking (MOT) in video streams. The method incorporates the prediction of physical locations of people into a tracking-by-detection paradigm. We predict the trajectories of people on an estimated ground plane and apply a learning-based network to extract the appearance features across frames. The method transforms the detected object locations from image space to an estimated ground space to refine the tracking trajectories. This transform space allows the objects detected from multi-view images to be associated under one coordinate system. Besides, the occluded pedestrians in image space can be well separated in a rectified ground plane where the motion models of the pedestrians are estimated. The effectiveness of this method is evaluated on different datasets by extensive comparisons with state-of-The-Art techniques. Experimental results show that the proposed method improves MOT tasks in terms of the number of identity switches (IDSW) and the fragmentations (Frag).

Original languageEnglish
Pages (from-to)137-143
Number of pages7
JournalISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Volume5
Issue number4
DOIs
Publication statusPublished - 2022
Event2022 24th ISPRS Congress on Imaging Today, Foreseeing Tomorrow, Commission IV - Nice, France
Duration: 6 Jun 202211 Jun 2022

Keywords

  • Data Association
  • Deep Features
  • Multiple Object Tracking
  • Tracking-by-Detection
  • Transform Space.

Fingerprint

Dive into the research topics of 'Multiple object tracking using a transform space'. Together they form a unique fingerprint.

Cite this