Image-based classification of sidescan sonar detections

PDF

Authors
  1. Fawcett, J.A.
Corporate Authors
Defence Research Establishment Atlantic, Dartmouth NS (CAN)
Abstract
In this report it is shown that the small images about target detections in sidescan sonar imagery can be used to classify the targets. In other words, the image itself is considered as a feature. This approach avoids the problems that traditional image-segmentation, feature-extraction and classification may entail. In particular, a situation where the background seabed imagery interferes with image-segmentation techniques is considered. The small images are analyzed using principal component decomposition, discriminate analysis and clustering. As well, a commercial software package is also used.

Il y a un résumé en français ici.

Keywords
Feature extraction;Image registration;Principal component analysis;Discriminant analysis;Clustering
Report Number
DREA-TM-2001-026 — Technical Memorandum
Date of publication
01 Apr 2001
Number of Pages
35
DSTKIM No
CA020018
CANDIS No
516682
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: