ME Special Seminar: Takeshi Hatanaka, "Cooperative Control & Estimation for Visual Sensor Networks"
Friday, November 15, 2013
DeWalt Seminar Room, 2164 Martin Hall
For More Information:
Mechanical Engineering Special Seminar
Cooperative Control & Estimation for Visual Sensor Networks
Takeshi Hatanaka, Ph.D.
Department of Mechanical & Control Engineering
Tokyo Institute of Technology, Japan
A visual sensor network is a network consisting of spatially distributed smart cameras with communication and computation capability. Typically, various decision-making and situation awareness processes in the current systems are conducted at a monitoring center by human operator(s), and partial distributed computing at each sensor is, if at all, done independently of the other sensors. However, since the volume of vision data tends to be larger than the other sensors, it is desired to distribute the entire process to each sensor while achieving total optimization through cooperation among sensors. In this talk, we will discuss how cooperative control and estimation techniques are installed into the networked vision systems. In the first part, we address cooperative estimation of 3D target motion. Here, we present a novel cooperative estimation mechanism by combining two of our previous works using passivity in rigid-body motion: vision-based 3D motion estimation and pose synchronization. Then, we analyze the averaging performance attained by the present estimation mechanism and demonstrate the effectiveness through experiments. In the second part, we investigate attitude control of vision sensors to achieve desirable visual coverage of a target space. For this purpose, we first formulate the visual coverage problem as an optimization problem on a matrix manifold and present a gradient descent control scheme. Then, we also present a density estimation algorithm including image processing and curve ﬁtting techniques by taking human motion monitoring as an example, and show the utility of the total process through simulation.
Takeshi Hatanaka received his B.Eng degree in informatics and mathematical science in 2002, and his M.Inf. and Ph.D. degrees in applied mathematics and physics in 2004 and 2007 respectively, all from Kyoto University, Japan. From 2006 to 2007, he was a research fellow of the Japan Society for the Promotion of Science at Kyoto University. He is currently an assistant professor in the Department of Mechanical and Control Engineering, Tokyo Institute of Technology, Japan. He received the Best Paper Award from SICE in 2009.
This Event is For: Graduate • Undergraduate • Faculty • Post-Docs • Alumni • Corporate