Mine Classification based on a Fuzzy Characterisation

Abstract : High resolution sonars provide high-quality acoustic images, allowing the classification of objects from their cast shadow. For a given ground mine except mine with radial symmetry, shadow appearance generally depends on the point of view. After a segmentation step performed on images acquired along a part of a circular trajectory of the sonar around the object, we can match and superimpose binary data. The resulting image displays a fuzzy shadow region whose pixels grey-levels depend on their successive localisation in the images of the sequence, i.e. if they belong or not to the shadow region. As an extension of feature extraction in the binary case, fuzzy geometry is a practical tool to describe fuzzy regions characterised by the degree of membership of each pixel to them. After a Principal Component Analysis applied to a set of fuzzy features, encouraging results have been achieved on simulated sonar images covering both classical and stealthy mines.
Complete list of metadatas

Cited literature [9 references]  Display  Hide  Download

Contributor : Isabelle Quidu <>
Submitted on : Wednesday, July 21, 2010 - 3:52:36 PM
Last modification on : Thursday, December 19, 2019 - 1:29:05 AM
Long-term archiving on: Friday, October 22, 2010 - 4:18:39 PM


Files produced by the author(s)


  • HAL Id : hal-00504850, version 1


Isabelle Quidu, Jean-Philippe Malkasse, Pierre Vilbé, Gilles Burel. Mine Classification based on a Fuzzy Characterisation. Undersea Defence Technology (UDT) Europe 2002, Jun 2002, La Spezia, Italy. ⟨hal-00504850⟩



Record views


Files downloads