Artur Loza

University of Bristol, Electrical & Electronic Engineering, Post-Doc

Followers

Following

Co-authors

Public Views

RESEARCH INTEREST:
Signal and image processing,
multimodal image fusion,
statistical image modelling,
video object tracking,
time-frequency signal processing,
statistical pattern recognition,
music technology.

less

Ilia Rodov

Bar-Ilan University

Fabio Cuzzolin

Oxford Brookes University

Roshan Chitrakar

Nepal College of Information Technology

Lev Manovich

Graduate Center of the City University of New York

Ilia Palaguta

Saint-Petersburg Stieglitz State Academy of Art and Design

Isad Saric

University of Sarajevo

barbara M . T . P . baert

KU Leuven

Bissera Pentcheva

Stanford University

PALIMOTE JUSTICE

RIVERS STATE POLYTECHNIC

Umit Unver

Yalova University

Interests

Uploads

Papers by Artur Loza

Advanced Study Institute Conference (NATO-ASI), Bulgaria

C.N.: The influence of multi-sensor video fusion on object tracking using a particle filter

This paper investigates howthe object tracking performance is affected by the fusion quality of v... more This paper investigates howthe object tracking performance is affected by the fusion quality of videos from visible (VIZ) and infrared (IR) surveillance cameras, as compared to tracking in single modality videos. The videos have been fused using the simple averaging, and various multiresolution techniques. Tracking has been accomplished by means of ap article filter using colour and edge cues. The highest tracking accuracyhas been obtained in IR sequences, whereas the VIZ video wasaffected by manyartifacts and showed the worst tracking performance. Among the fused videos, the complexw avelet and the averaging techniques, offered the best tracking performance, comparable to that of IR. Thus, of all the methods investigated,the fused videos, containing complementary contextual information from both single modality input videos, are the best source for further analysis by ahuman observer or acomputer program. 1I ntroduction Recently thereh as been an increased interest in object tracking in video sequences supplied by as ingle camerao ranetwork of cameras [FMS + 05, HBC + 05, PVB04, BK05, YSS03]. Reliable tracking methods are of crucial importance in manys urveillance systems as theyenable human operators to remotely monitor activity across areas of interest, enhance situation awareness and help the surveillance analyst with the decision-making process. Multiple-sensor systems can provide surveillance coverage across aw ide area and under different conditions.T he complementarity of the information supplied by different sensors is valuable for solving different kinds of problems, such as detection, recognition, tracking and situation assessment, to name butafew, and ought to be fused in order to servemore efficiently these purposes. Data fusion comprises groups of methods for merging data from multiple sensors. In this paper,w ei nvestigate howt he fusion of video sequences from visible (VIZ) and infrared (IR) cameras influences the process of object tracking. We compare the tracking results from the fused sequences with the results obtained from separate IR andVIZ videos. The remaining part of this paper is organised as follows. Section 2presents the methodology used to assess the particle filter tracking performance in multi-sensor real video data. Conclusions and future plans are giveninSection 3.

Artur Loza

Related Authors

Uploads

Papers by Artur Loza

Log In