Feature Extraction and Classification Methods for Ultra-Sonic and Radar Mine Detection

Research output: Chapter in Book/Report/Conference proceedingConference paper

Abstract

In this paper the problem of detecting buried anti-personnel mines is tackled in the broader context of a general classication problem: determining the likelihood of an unknown pattern (feature vector), extracted from the impulse radar or ultra sound signals, being part of a mine. In our approach two scenarios are considered. In a rst scenario we use a single classier with a mixed measurement vector, while in a second scenario a strategy for combining several classiers, using distinct features is explained.
Original languageEnglish
Title of host publicationCESA'98, 2nd IEEE- IMACS Int. Multiconference on Computational Engineering in Systems Applications; Tunis, Tunisia, April 1-4,1998.
PublisherSecond Int. Multiconference on Computational Engineering in Systems Applications (CESA 98), pp. 82 - 87, Tunis, Tunisia.
Pages82-87
Number of pages6
Publication statusPublished - 1 Apr 1998

Bibliographical note

Second Int. Multiconference on Computational Engineering in Systems Applications (CESA 98), pp. 82 - 87, Tunis, Tunisia.

Fingerprint

Dive into the research topics of 'Feature Extraction and Classification Methods for Ultra-Sonic and Radar Mine Detection'. Together they form a unique fingerprint.

Cite this