This paper proposes a framework in which Lagrangian Particle Dynamics is used for the segmentation of high density crowd flows and detection of flow instabilities. For this purpos...
Problem of segmenting individual humans in crowded situations from stationary video camera sequences is exacerbated by object inter-occlusion. We pose this problem as a “model-b...
Segmentation and tracking of multiple humans in crowded situations is made difficult by interobject occlusion. We propose a model-based approach to interpret the image observations...
Given a collection of images of a static scene taken by many different people, we identify and segment interesting objects. To solve this problem, we use the distribution of images...
This paper presents a fast, accurate, and novel method for the problem of estimating the number of humans and their positions from background differenced images obtained from a si...
Lan Dong, Vasu Parameswaran, Visvanathan Ramesh, I...