INTRODUCTION OF THE SIGNATURE CLASSIFICATION DEVELOPMENT SYSTEM

PDF

Authors
  1. Brown, L.M.
  2. Lin, J.S.
Corporate Authors
Naval Surface Warfare Center, Annapolis MD (US) Carderock Div;Defence Research Establishment Atlantic, Dartmouth NS (CAN)
Abstract
The Carderock Division of the Naval Surface Warfare Center and the Applied Physics Laboratory of the Johns Hopkins University are currently developing a Signature Classification Development System (SCDS) to transfer signature classification technology to NDE inspection equipment in the field. The SCDS is a PC-based software tool-kit designed to aid in the analysis of NDE signatures and to support the development of classification algorithms. Three modules are integral to the Signature Classification Development Systems: (1) the digital signal processing module (DSP), the classification module, and the code generation module. The DSP module supports user-written C routines to facilitate signal processing, definition of templates, assignment of signature gates, and generation of feature vectors. The SCDS classification module incorporates an artificial neural network and both parametric and non-parametric statistical techniques for classification of the NDE signatures. After a satisfactory classification algorithm has been developed and tested, the SCDS can generate run-time source code for integration of the processing and classification algorithm into NDE field equipment or laboratory systems. The Signature Classification Development System is an on-going program sponsored by the Office of Naval Research. This paper briefly describes the current capabilities of the SCDS and outlines the plans for future enhancements.
Report Number
DREA-NAMT-95-2-PAP-2 — CONTAINED IN 99-00779
Date of publication
30 May 1995
Number of Pages
8
DSTKIM No
99-00785
CANDIS No
510349
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: