NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
Testing Saliency Parameters for Automatic Target RecognitionA bottom-up visual attention model (the saliency model) is tested to enhance the performance of Automated Target Recognition (ATR). JPL has developed an ATR system that identifies regions of interest (ROI) using a trained OT-MACH filter, and then classifies potential targets as true- or false-positives using machine-learning techniques. In this project, saliency is used as a pre-processing step to reduce the space for performing OT-MACH filtering. Saliency parameters, such as output level and orientation weight, are tuned to detect known target features. Preliminary results are promising and future work entails a rigrous and parameter-based search to gain maximum insight about this method.
Document ID
20150005545
Acquisition Source
Jet Propulsion Laboratory
Document Type
Other
External Source(s)
Authors
Pandya, Sagar
(University of Southern California Los Angeles, CA, United States)
Date Acquired
April 10, 2015
Publication Date
August 1, 2012
Subject Category
Earth Resources And Remote Sensing
Distribution Limits
Public
Copyright
Other
Keywords
regions of interest (ROI)
Automatic target recognition (ATR
machine-learning algorith

Available Downloads

There are no available downloads for this record.
No Preview Available