Object joint detection and tracking using adaptive multiple motion models

Authors:
Zhijie Wang;Mohamed Ben Salah;Hong Zhang
Affiliations:
Department of Computing Science, University of Alberta, Edmonton, Canada;Department of Computing Science, University of Alberta, Edmonton, Canada;Department of Computing Science, University of Alberta, Edmonton, Canada
Venue:
The Visual Computer: International Journal of Computer Graphics
Year:
2014

Citing 9
Cited 0

Estimation with Applications to Tracking and Navigation

Estimation with Applications to Tracking and Navigation
Tracking a manoeuvring target using angle-only measurements: algorithms and performance

Signal Processing
Recognizing Temporal Trajectories Using the Condensation Algorithm

FG '98 Proceedings of the 3rd. International Conference on Face & Gesture Recognition
A Mixed-State Condensation Tracker with Automatic Model-Switching

ICCV '98 Proceedings of the Sixth International Conference on Computer Vision
A particle filter for joint detection and tracking of color objects

Image and Vision Computing
Bearings-only tracking of manoeuvring targets using particle filters

EURASIP Journal on Applied Signal Processing
Large Lump Detection Using a Particle Filter of Hybrid State Variable

ICAPR '09 Proceedings of the 2009 Seventh International Conference on Advances in Pattern Recognition
Learning Generative Models for Multi-Activity Body Pose Estimation

International Journal of Computer Vision
Artificial Intelligence: A Modern Approach

Artificial Intelligence: A Modern Approach

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper deals with the problem of detecting objects that may switch between different motion models. In order to accurately detect these moving objects taking into account possible changing motion models, we propose an adaptive multi-motion model in the joint detection and tracking (JDT) framework. The proposed technique differs from the existing JDT-based methods mainly in two ways. First we express the solution in the JDT framework via a formulation in the multiple motion model setting. Second, we introduce a new motion model prediction function which exploits the correlation between the motion model and object kinematic state. Experiments on both synthetic and real videos demonstrate that the JDT method employing the proposed adaptive multi-motion model can detect objects more accurately than the existing peer methods when objects change their motion models.