Global trajectory reconstruction from distributed visual sensors

Kayumbi, Gabin; Anjum, Nadeem; Cavallaro, Andrea

doi:10.1109/icdsc.2008.4635697

Cited by 12 publications

(7 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We compare the performance of the proposed approach with standard Dynamic Time Warpping (DTW) [5] (M1) and two state-of-the-art approaches presented in [4] (M2) and [7] (M3) in terms of P and R for both sequences. The results are compiled in Table 1.…”

Section: Resultsmentioning

confidence: 99%

“…Unlike supervised techniques, unsupervised techniques do not require training samples or manual selection of the parameters. Recent unsupervised target as- sociation algorithms are presented by Kayumbi et al [4] and Sheikh et al [7]. The rest of this section provide the details of these techniques.…”

Section: Prior Workmentioning

confidence: 99%

“…Kayumbi et al [4] establish correspondence between cameras and virtual ground-plane. Trajectory association is done on the ground-plane using shape and length along with temporal information.…”

Section: Prior Workmentioning

confidence: 99%

“…We assume that a pair of trajectories from different cameras has to be similar both in time and space for the association and fusion. In [4], the shape (β i n,G ), which is approximated by polynomial coefficients, and length (d i n,G ) are used to find the similarity. However, these features are not generalized enough to handle variety of trajectories.…”

Section: Trajectory Associationmentioning

confidence: 99%

“…Existing works perform association either on image-plane [7] or on ground-plane [4]. As image-plane trajectories are heavily affected by the perspective deformations, which cause inaccurate associations especially if the trajectories are far from cameras.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Trajectory Association and Fusion across Partially Overlapping Cameras

Anjum

Cavallaro

2009

2009 Sixth IEEE International Conference on Advanced Video and Signal Based Surveillance

Self Cite

View full text Add to dashboard Cite

We present a novel unsupervised inter-camera trajectory correspondence algorithm that does not require prior knowledge of the camera placement. The approach consists of three steps, namely association, fusion and linkage. For association, local trajectory pairs corresponding to the same physical object are estimated using multiple spatio-temporal features on a common ground-plane. To disambiguate spurious associations, we employ a hybrid approach that utilizes the matching results on the imageand ground-plane. The trajectory segments after association are fused by adaptive averaging. Finally, linkage integrates segments and generates a single trajectory of an object across the entire observed area. We evaluated the performance of the proposed approach on a simulated and two real scenarios with simultaneous moving objects observed by multiple cameras and compared it with state-of-the-art algorithms. Convincing results are observed in favor of the proposed approach.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Prior Workmentioning

confidence: 99%

Section: Prior Workmentioning

confidence: 99%

Section: Trajectory Associationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations