CONSTRUCTIVE NEURAL NETWORK FOR LANDMINE CLASSIFICATION USING ULTRA WIDEBAND GPR
In this paper, constructive neural network for landmine classification using ultra wideband (UWB) Ground Penetrating Radar (GPR) is presented. GPR echo signal is composed of three parts: ground bounce, clutter and target echo signal, the target echo signal is deteriorated by the clutter. Firstly WP-based preprocessing algorithm is used to ground bounce removal and clutter reduction and feature extraction of GPR echo signal. Then wrapper based approach is adopted to feature subset selection of GPR echo signal using genetic algorithm(GA) in conjunction with constructive neural network learning algorithm, and at the meanwhile, the result of classification of landmine is obtained. Experiment result based on GPR measured data shows that the feasibility and advantage of the presented algorithm.
UWB GPR WP-based preprocessing algorithm constructive neural network landmine classification
HUI-LIN ZHOU WEI-PING WANG YU-HAO WANG
School of Information Engineering, Nanchang University, Jiangxi, China
国际会议
2008 International Conference on Machine Learning and Cybernetics(2008机器学习与控制论国际会议)
昆明
英文
1197-1201
2008-07-12（万方平台首次上网日期，不代表论文的发表时间）