Border and Coastal Surveillance Sound Target Classification Based on Optimum Wavelet Packet Decomposition

Yan Wang; Zhi Li

doi:10.4028/www.scientific.net/AMR.225-226.725

Paper Titles

Cloud Computing Security Issues and Mechanisms
p.706

Study on Design and Development of Mechanical Engineering Network Courses
p.710

Research on Moisture Damage Research in Hunan Province Expressway Pavement
p.714

Investigation on Knowledge Work Process and its Characteristics Based on Human-Information System (H-IS) Interaction
p.720

Border and Coastal Surveillance Sound Target Classification Based on Optimum Wavelet Packet Decomposition
p.725

Research on Enterprise Application Integration Based on Service Oriented and Workflow
p.729

Utility-Based Optimal Multi-Flow Rate Allocation for Concurrent Multipath Transport among Cooperative Terminals
p.734

Development Strategy for Requirement of ICTs in Engineering Learning of Comprehensive Regional Higher Education Institutes of PRC
p.739

A Brief Analysis on the Construction of Banking Office Automation System
p.743

HomeAdvanced Materials ResearchAdvanced Materials Research Vols. 225-226Border and Coastal Surveillance Sound Target...

Border and Coastal Surveillance Sound Target Classification Based on Optimum Wavelet Packet Decomposition

Abstract:

The present work contributes to the field of border and coastal surveillance sound target classification. A new feature extraction method is proposed based on the optimum wavelet packet decomposition (OWPD). According to the frequency characteristic of border and coastal surveillance sound signals, each signal is decomposed by selective multi-scale wavelet packet decomposition (WPD) and the OWPD tree is obtained. From their high dimension OWPD coefficients, we build the meaningful and compact energy feature vectors, then use them as the input vectors of the BP neural network to classify the border and coastal surveillance sound types. Extensive experimental results show that the classification efficiency is up to 94% using this feature extraction method, improved 6% compared with the method based on WPD.