Avinash Kumar
Phone: (217) 244-4392
Address:
University of Illinois
1520 Beckman Institute
405 N. Mathews Ave.
Urbana, IL 61801
Biography:
Avinash Kumar received his B.Tech degree in Computer Science from International Institute of Information Technology at Hyderabad (IIIT-H) in August 2005. Thereafter, he obtained MS by Research degree, specializing in Computer Vision at CVIT under the supervision of Dr. C.V. Jawahar in December 2007. Beginning January 2008, he has been a Ph.D. student in the Electrical and Computer Engineering Department at the University of Illinois Urbana-Champaign and is advised by Prof. Narendra Ahuja. His research focuses on optical camera designs for accurate scene depth estimation and application of computer vision to solve real world problem of train monitoring.
He is the co-recipient of T.A. Stewart-Dyer and F. Trevithick Prize (2007) awarded by the Institution of Mechanical Engineers for his work on the use of computer vision techniques for increasing the fuel efficiency of intermodal freight trains.
Research Description
Depth From Focus: 3D scene depth can be estimated using a number of cues. In this work we use the combination of focus and defocus information in a set of multifocus images to estimate the scene depth. The system generates the scene depth along with an omnifocus image in which all objects are in focus irrespective of their depth.
Publications: None
Optimal Camera Parameters : In this work we find the optimal camera calibration parameters for a special camera used for capturing panoramic multifocus images. The obtained parameters for this camera ensure that minimum number of such images are captured such that they encode the maximum information about the scene which is necessary and sufficient for accurate estimation of scene depth and for corresponding omnifocus image generation.
Publications : None
Intermodal Freight Train Analysis System: From a video of freight train, we apply computer vision techniques for removing background and estimating velocity of a train. This leads to automatic train inspection and calculation of gap lengths between consecutive loads. This information is useful for designing optimal loading patterns thus saving on a non-renewable source of energy i.e. fuel.
Publications:
A. Kumar, N. Ahuja, J. Hart, U.K. Visesh, P.J. Narayanan, C.V. Jawahar, A Vision System for Monitoring Intermodal Freight Trains,
Workshop on Applications of Computer Vision (WACV), Austin, TX, February 2007. [Full Text]
Lai, Y-C., C.P.L. Barkan, J. Drapa, N. Ahuja, J.M. Hart, P.J. Narayanan, C.V. Jawahar, A. Kumar, L. Milhon and M.P. Stehly 2007.
Machine-vision analysis of the energy efficiency of intermodal freight trains. Journal of Rail and Rapid Transit 221: 353-364. [Full Text] (T.A. Stewart-Dyer and F. Trevithick Prize Winner)
Q.Kong, A.Kumar, N.Ahuja and Y.Liu, Robust Segmentation of Freight Containers in Train Monitoring Videos, Workshop on Applications of Computer Vision(WACV), Snowbird, Utah, December 2009. (Oral Acceptance Rate ~ 12%)