By Topic

Interest Segmentation of Large Area Spectral Imagery for Analyst Assistance

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Ariel Schlamm ; Center for Imaging Science, Rochester Institute of Technology, Rochester, NY, USA ; David Messinger ; William Basener

Widely used methods of spectral clustering, target, and anomaly detection when applied to spectral imagery provide less than desirable results across sensor type, scene content, spectral and spatial resolutions due to the complex nature of the data. This results in a large burden placed on the analyst in terms of the amount of data needed to be processed and the ability to discern the difference between “interesting” and “uninteresting” regions in the imagery. For this research, a variety of data driven algorithms for spectral image analysis are applied to spatial tiles of a large area scene. A feature map is created by assigning a metric determined for each algorithm result to each spatial tile. The feature maps are organized into a tiled, multi-band feature image. Two visualization methods introduced here provide a detection map which can cue image analysts to visually inspect locations within a large area scene with a high likelihood of containing interest. Unsupervised classification is applied to this feature image such that the image is divided into segments representing either “interesting” or “not interesting” content with the tile. False-color visualization of three independent metrics is also presented as a way to indicate the type and strength of the amount of interest within a tile.

Published in:

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  (Volume:5 ,  Issue: 2 )