摘要
鉴于声目标识别在民用和军用方面的广泛应用,对声目标进行识别已成为模式识别领域中的研究热点之一。使用小波分析实现对声信号的消噪处理,采用小波包能量特征提取方法获得目标声信号的特征向量,分类器设计采用支持向量机的模式识别方法,对声目标识别系统的组成、功能结构和框架结构进行了设计,同时介绍了系统数据库设计的要求和方法,并给出了基于VC与MATLAB联合编程的声目标识别系统中的主要算法及实现程序。仿真实验表明,系统能有效地对声目标进行分类识别。
Due to the wide application of acoustic tea＇get recognition in civil and military aspects, it has been a hotspot of study for pattern recognition. Designed the composition, function structure, framework structure of the acoustic target recognition system based on restrain noise and feature extraction with wavelet and wavelet packet, support vector machine（SVM） classifier. At the same time, the design demands and methods of database is proposed and the main algorithm and program based on VC and MATLAB of acoustic target recognition system is given. The simulation experiments show that the system designed by this article can effectively classify acoustic targets.
出处
《计算机技术与发展》
2009年第9期207-210,共4页
Computer Technology and Development
基金
云南省自然科学基金项目(2005F0031Q)
云南省教育基金(07C41491)
