Distributed Camera Networks
Camera networks have become increasingly important in recent years. This is primarily attributed to their increasing adoption and usage around the world. This project works on data collected from a distributed camera network comprising of non-overlapping camera views. The first problem we addressed was the design of a scale aware optical flow descriptor for understanding motion patterns occurring in different cameras. This descriptor is lightweight, making it easy to transmit and scale across streaming dat. Subsequently, we addressed the problem of answering queries of where a certain person of interest appeared in any camera view by modeling the delay between cameras, in conjunction with the optical flow descriptor. This led to the design of the first query answering system in a distributed camera network with non-overlapping views.
Relevant Publications
Graph-based Topic-focused Retrieval in Distributed Camera Network
Jiejun Xu, Vignesh Jagadeesh, Zefeng Ni, Santhosh KumarSunderrajan, B.S. Manjunath
IEEE Transactions on Multimedia, TMI 2014
Jiejun Xu, Vignesh Jagadeesh, Zefeng Ni, Santhosh KumarSunderrajan, B.S. Manjunath
IEEE Transactions on Multimedia, TMI 2014