Object Tracking Using Grayscale Appearance Models and Swarm Based Particle Filter

Authors:
Bogdan Kwolek
Affiliations:
Rzeszów University of Technology, Rzeszów, Poland
Venue:
HAIS '08 Proceedings of the 3rd international workshop on Hybrid Artificial Intelligence Systems
Year:
2008

Citing 8
Cited 2

Color-Based Probabilistic Tracking

ECCV '02 Proceedings of the 7th European Conference on Computer Vision-Part I
Color-Based Tracking of Heads and Other Mobile Objects at Video Frame Rates

CVPR '97 Proceedings of the 1997 Conference on Computer Vision and Pattern Recognition (CVPR '97)
Unsupervised Improvement of Visual Detectors using Co-Training

ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
Integral Histogram: A Fast Way To Extract Histograms in Cartesian Spaces

CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Kernel Particle Filter for Real-Time 3D Body Tracking in Monocular Color Images

FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
Learning-based object tracking using boosted features and appearance-adaptive models

ACIVS'07 Proceedings of the 9th international conference on Advanced concepts for intelligent vision systems
Learning compositional categorization models

ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part III
Robust online appearance models for visual tracking

IEEE Transactions on Pattern Analysis and Machine Intelligence

Particle Swarm Optimization Based Object Tracking

Fundamenta Informaticae - Swarm Intelligence
Efficient multi-feature PSO for fast gray level object-tracking

Applied Soft Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose a hybrid tracking algorithm consisting of two trackers built on grayscale appearance models. In a first tracker we employ an object template that consists of several grayscale image patches. Every patch votes for the possible positions of the object undergoing tracking. A grayscale appearance model that is learned on-line is used in a supplementing tracker. A particle swarm optimization algorithm is utilized to shift particles toward more promising regions in the probability density function. Experimental results show that the hybrid tracker outperforms each of the trackers.