南京大学学报(自然科学版) ›› 2015, Vol. 51 ›› Issue (7): 16.
孙馨喆1,2,文立1,2,杨武夷1,2*,张宇1,2
Sun Xinzhe1,2, Wen li1,2, Yang Wuyi1,2*, Zhang Yu1,2
摘要: 人工检测宽吻海豚声通讯信号效率低下。本文提出了一种基于信号时频图像处理的方法,旨在实现对宽吻海豚声通讯信号的高效自动检测。该方法首先对采集到的音频信号进行分帧,计算每帧信号的时频图;对时频图进行中值滤波后利用自适应局部阈值法提取时频图中通讯信号的轮廓;对提取的轮廓进行连通域分析,根据各连通域的能量和持续时间过滤噪声连通域;合并邻近的连通域后根据连通域的位置判断海豚通讯信号在音频信号中的起止时刻。实验结果显示本文的检测方法准确率达到了95%,该方法为海豚声信号的自动观测及其生物学行为的研究提供了一定的技术基础。
[1] Perrin W F, Bernd W, Thewissen J G M. Encyclopedia of Marine Mammals. SanDiego: Academic Press, 2002, 122127. [2] Lilly J C. Sonic-ultrasonic emissions of the bottlenose dolphin. Whales, dolphins and porpoises, 1966, 165: 503-507. [3] Whitlow W L Au. The Sonar of Dolphins. New York: Springer-Verlag, 1993. [4] Mellinger D K, Martin S W, Morrissey S W, et al. A method for detecting whistles, moans,and other frequency contour sounds. Journal of the Acoustical Society of America, 2011, 129(6): 4055-4061. [5] Parada P P, Cardenal-Lopez A. Using Gaussian mixture models to detect and classify dolphin whistles andpulses. Journal of the Acoustical Society of America, 2014, 135(6): 3371-3380. [6] Lin T H, Yu H Y, Chen C F, et al. Automatic detection and classification of cetacean tonal sounds from a long-term marine observatory. Proceedings of Symposium on Underwater Technology. Tokyo: IEEE Press, 2013:1-6. [7] Gillespie D, Caillat M, Gordon J, et al. Automatic detection and classification of odontocete Whistles. Journal of the Acoustical Society of America, 2013, 134(3): 2427-2437. [8] Roch M A, Brandes T S, Patel B. Automated extraction of odontocete whistle contours. Journal of the Acoustical Society of America, 2011, 130: 2212-2223. |
No related articles found! |
|