A Unified Framework for Tracking through Occlusions and across Sensor Gaps

Kaucic, R.; Perera, A. G. Amitha; Brooksby, G.; Kaufhold, J.; Hoogs, Anthony

doi:10.1109/cvpr.2005.53

Cited by 93 publications

(80 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Modern systems can track through long and challenging sequences with high precision. To this end, researchers have focused on improving the appearance model [10,5], the object detector [2,7,9,22], and/or the optimization strategy [14,16,23]. Others have developed approaches specifically for crowded scenes [1,6,24].…”

Section: Related Workmentioning

confidence: 99%

You'll never walk alone: Modeling social behavior for multi-target tracking

Pellegrini

Ess

Schindler

et al. 2009

2009 IEEE 12th International Conference on Computer Vision

1,199

1,095

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

You'll never walk alone: Modeling social behavior for multi-target tracking

Pellegrini

Ess

Schindler

et al. 2009

2009 IEEE 12th International Conference on Computer Vision

1,199

1,095

View full text Add to dashboard Cite

“…When objects are detected in cameras separately, reduction in detection regions results in the loss of overlap between two cameras. While methods for matching objects across non-overlapping cameras exist [1,11,12,6], low resolution and single channel data disallow the use of appearance models for object hand over, and reacquisition based on motion alone is ambiguous. The increased gap between cameras arising from detection adds further challenge to a data already characterized by high density of objects and low sampling rate of video.…”

mentioning

confidence: 99%

Detection and Tracking of Large Number of Targets in Wide Area Surveillance

Reilly

Idrees

Shah

2010

Lecture Notes in Computer Science

143

147

View full text Add to dashboard Cite

Abstract. In this paper, we tackle the problem of object detection and tracking in a new and challenging domain of wide area surveillance. This problem poses several challenges: large camera motion, strong parallax, large number of moving objects, small number of pixels on target, single channel data and low framerate of video. We propose a method that overcomes these challenges and evaluate it on CLIF dataset. We use median background modeling which requires few frames to obtain a workable model. We remove false detections due to parallax and registration errors using gradient information of the background image. In order to keep complexity of the tracking problem manageable, we divide the scene into grid cells, solve the tracking problem optimally within each cell using bipartite graph matching and then link tracks across cells. Besides tractability, grid cells allow us to define a set of local scene constraints such as road orientation and object context. We use these constraints as part of cost function to solve the tracking problem which allows us to track fast-moving objects in low framerate videos. In addition to that, we manually generated groundtruth for four sequences and performed quantitative evaluation of the proposed algorithm.

show abstract

“…This danger can be reduced by optimizing data assignment and considering information over several time steps, as in MHT [18] and JPDAF [6]. However, task complexity limits previous optimization approaches to consider either only few time steps [18] or only single trajectories over longer time windows [2,11,23]. In contrast, our approach simultaneously optimizes detection and trajectory estimation for multiple interacting objects and over long time windows by operating in a hypothesis selection framework.…”

Section: Related Workmentioning

confidence: 99%

“…Our approach implements a feedback loop, which passes on predicted object locations as a prior to influence detection, while at the same time choosing between and reevaluating trajectory hypotheses in the light of new evidence. In contrast to previous approaches, which optimize individual trajectories in a temporal window [2,23] or over sensor gaps [11], our approach tries to find a globally optimal combined solution for all detections and trajectories, while incorporating physical constraints such that no two objects can occupy the same physical space, nor explain the same image pixels at the same time. The task complexity is reduced by only selecting between a limited set of plausible hypotheses, which makes the approach computationally feasible.…”

Section: Introductionmentioning

confidence: 99%