Stationary objects in multiple object tracking

Authors:
Sadiye Guler;Jason A. Silverstein;Ian H. Pushee
Affiliations:
intuVision, Inc. 10 Tower Office Park Ste.200 Woburn, Massachusetts 01801, USA;intuVision, Inc. 10 Tower Office Park Ste.200 Woburn, Massachusetts 01801, USA;intuVision, Inc. 10 Tower Office Park Ste.200 Woburn, Massachusetts 01801, USA
Venue:
AVSS '07 Proceedings of the 2007 IEEE Conference on Advanced Video and Signal Based Surveillance
Year:
2007

Citing 0
Cited 5

3D Neural Model-Based Stopped Object Detection

ICIAP '09 Proceedings of the 15th International Conference on Image Analysis and Processing
Comparative Evaluation of Stationary Foreground Object Detection Algorithms Based on Background Subtraction Techniques

AVSS '09 Proceedings of the 2009 Sixth IEEE International Conference on Advanced Video and Signal Based Surveillance
Real-time illegal parking detection in outdoor environments using 1-D transformation

IEEE Transactions on Circuits and Systems for Video Technology
Localized detection of abandoned luggage

EURASIP Journal on Advances in Signal Processing - Special issue on video analysis for human behavior understanding
STV-based video feature processing for action recognition

Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents an approach to detect stationary foreground objects in naturally busy surveillance video scenes with several moving objects. Our approach is inspired by human’s visual cognition processes and builds upon a multi-tier video tracking paradigm with main layers being the spatially based “peripheral tracking” loosely corresponding to the peripheral vision and the object based “vision tunnels” for focused attention and analysis of tracked objects. Humans allocate their attention to different aspects of objects and scenes based on a defined task. In our model, a specific processing layer corresponding to allocation of attention is used for detection of objects that become stationary. The static object layer, a natural extension of this framework, detects and maintains the stationary foreground objects by using the moving object and scene information from Peripheral Tracker and the Scene Description layers. Simple event detection modules then use the enduring stationary objects to determine events such as Parked Vehicles or Abandoned Bags.