Continuous Multi-Views Tracking using Tensor Voting
نویسندگان
چکیده
This paper presents a new approach for continuous tracking of moving objects observed by multiple fixed cameras. The continuous tracking of moving objects in each view is realized using a Tensor Voting based approach. We infer objects trajectories by performing a perceptual grouping in 2D+t using Tensor Voting. Also, a multi-scale approach bridging gaps in object trajectories is presented. The trajectories obtained from the multiple cameras are registered in space and time allowing a synchronization of the video streams and a continuous tracking of objects across multiple views. We demonstrate the performance of the system on several real video surveillance sequences.
منابع مشابه
Continuous Multi-View Tracking using Tensor Voting
We present a novel approach for continuous tracking of moving objects observed by multiple fixed cameras. The continuous tracking of moving objects in each view is realized using a Tensor Voting based approach. We infer objects trajectories by performing a perceptual grouping in 2D+t. Gaps are bridged using a multi-scale approach in object trajectories. The trajectories obtained from the multip...
متن کاملMultiple People Tracking Using Non-iterative Tensor Voting Method
Human tracking is an emerging research field in computer vision and video surveillance. Lightning changes and occlusion are the main challenges faced during tracking. This paper focuses on people tracking both in indoor and outdoor scenes. Our method is based on color modeling using tensor voting framework, which can handle these challenges. The lightning conditions are controlled by homomorphi...
متن کاملPerceptual Grouping Based on Iterative Multi-scale Tensor Voting
We propose a new approach for perceptual grouping of oriented segments in highly cluttered images based on tensor voting. Segments are represented as second-order tensors and communicate with each other through a voting scheme that incorporates the Gestalt principles of visual perception. An iterative scheme has been devised which removes noise segments in a conservative way using multi-scale a...
متن کاملZoom-Invariant Tracking using Points and Lines in Affine Views - An Application of the Affine Multifocal Tensors
This paper describes methods for tracking using point and line features in affine views to provide fundamental invariance to changes of focal length. It first demonstrates how an earlier method of transfer-based tracking using spatio-temporal matching of point features in a stereo active head is indeed zoom-invariant. In order to also make use of lines the paper then illustrates how the affine ...
متن کاملA Closed-Form Solution to Tensor Voting: Theory and Applications
We prove a closed-form solution to tensor voting (CFTV): given a point set in any dimensions, our closed-form solution provides an exact, continuous and efficient algorithm for computing a structure-aware tensor that simultaneously achieves salient structure detection and outlier attenuation. Using CFTV, we prove the convergence of tensor voting on a Markov random field (MRF), thus termed as MR...
متن کامل