Multi-Target Detection and Tracking Over Multi-Camera Surveillance System
People
Ching{Chun Huang and Sheng{Jyh Wang
System Flow of the proposed algorith.
Abstract
With the increasing demands for security and safety, multi-camera video surveillance systems are gaining popularity. To assist users in efficiently summarizing information from multiple cameras, we aim to design a multi-camera system that can fuse information from a set of calibrated cameras to perform detection and tracking of multiple targets. To build such a system, we focus on three major issues: (1) this system shall provide reliable data fusion from multiple cameras; (2) this system can simultaneously detect and track multiple moving targets; and (3) the computational complexity of this system shall be affordable.
Several testing sequences and tracking results are provided for references. OVVV synthesized video 1 : Video sequence , Tacking results , and Combined results OVVV synthesized video 2 : Video sequence , Tacking results , and Combined results Real video in lab 1 : Video sequence, Background subtraction, Tacking results, and Combined results Real video in lab 2 : Video sequence, Background subtraction, Tacking results, and Combined results
Publication
Ching-Chun Huang, Sheng-Jyh Wang "A Monte Carlo Based Framework for Multi-Target Detection and Tracking Over Multi-Camera Surveillance System", accepted by the 10th European Conference on Computer Vision (ECCV) Workshop on Multi-camera and Multi-modal Sensor Fusion Algorithms and Applications, Marseille, France, October 12-18, 2008. NSC95-2221-E-009-106-MY2.