摘要
针对海豚哨声信号自动检测的问题,提出一种基于分形维数的自适应阈值海豚哨声信号检测方法。对待检测声信号计算盒分形维数,根据得到的盒分形维数特征值,通过模糊C均值聚类自适应确定检测阈值,实现海豚哨声信号的自动检测。文中对水池录制的海豚声信号进行了数据分析,利用哨声信号盒分形维数对哨声信号段与非信号段进行检测,并与基于谱熵的方法进行对比,获得了较高的检测率以及较低的虚警率,可以适用于海豚哨声信号的自动检测与分割。
Aiming at the problem of automatic detection of dolphin whistle signal,an adaptive threshold detection method based on fractal dimension was proposed.The detection threshold was determined by fuzzy C-means clustering based on the characteristic value of box fractal dimension,and the automatic detection of dolphin whistle signal was realized.The fractal dimension of whistle signal box is used to detect the whistle signal segment and non-signal segment,and compared with the method based on spectral entropy,higher detection rate and lower error detection rate are obtained,which can be applied to automatic detection and segmentation of dolphin whistle signal.
作者
张晓伟
尹力
薛山花
张春华
ZHANG Xiaowei;YIN Li;XUE Shanhua;ZHANG Chunhua(Institute of Acoustics,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China;Key Laboratory of Science and Technology on Advanced Underwater Acoustic Signal Processing,Chinese Academy of Sciences,Beijing 100190,China)
出处
《应用声学》
CSCD
北大核心
2023年第2期237-242,共6页
Journal of Applied Acoustics
基金
科技部重点研发项目(2018YFC1405904)。
关键词
海豚声信号
哨声信号
盒分形维数
模糊C均值聚类
谱熵
Dolphin signal
Whistle signal
Box fractal dimension
Fuzzy C-means clustering
Spectral entropy