Home -> Projects ->Flow-Based Background Estimation

Member Login

Jump to: Publications

Description

In this work, we address the detection of vehicles in a video stream obtained from a moving airborne platform. We propose a Bayesian framework for estimating dense optical flow over time that explicitly estimates a persistent model of background appearance. We use an approach that integrates dense optic flow over time to maintain a model of background appearance and a foreground occlusion layer mask.

Demos

Result Video 1: Smalls cars Download: avi [8.7MB]

Result Video 2: Occlusion and Shadows Download: avi [25MB]

Result Video 3: Occlusion and Motion Parallax Download: avi [23MB]

Result Video 4: Small cars Download: avi [19MB]

Result Video 5: Thermal sequence with changing camera gain Download: avi [12MB]

Publications