"Onboard real-time processing is highly desirable in airborne detection applications. As the data processing involved here is computationally expensive, typically high power multi-rack system is required to achieve real-time detection. Use of such hardware is often not feasible in an airborne application due to space, power and weight constraints. In this thesis, different computing architectures suitable for real-time onboard processing are compared in terms of computation speed, watts/GFLOPS, and weight/GFLOPS. A brief description of the modem GPU architecture and the current GPGPU technology is provided. GPU implementations of some computationally expensive image processing applications such as band-to-band registration for Multi Spectral Imagery (MSI), image registration, and multiband RX anomaly detector are described. The sub-problems that can be efficiently parallelized on the Single Instruction Multiple Data (SIMD) architecture of the GPU are identified and the results are compared to those obtained using Matlab and Intel Performance Primitives (IPP) on multicore Intel processors"--Abstract, page iii.
Agarwal, Sanjeev, 1971-
Moss, Randy Hays, 1953-
Madria, Sanjay Kumar
Electrical and Computer Engineering
M.S. in Computer Engineering
Missouri University of Science and Technology
viii, 42 pages
© 2010 Dheeraj Singiresu, All rights reserved.
Thesis - Restricted Access
Library of Congress Subject Headings
Graphics processing units
Image processing -- Computer simulation
High performance computing
Print OCLC #
Electronic OCLC #
Link to Catalog Record
Electronic access to the full-text of this document is restricted to Missouri S&T users. Otherwise, request this publication directly from Missouri S&T Library or contact your local library.http://laurel.lso.missouri.edu:80/record=b10158443~S5
Singiresu, Dheeraj, "GPU based processing for airborne detection" (2010). Masters Theses. 4496.
Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.