Automated Monitoring And Analysis Of Marine Mammal Vocalizations In Coastal Habitats

2018 OCEANS - MTS/IEEE KOBE TECHNO-OCEANS (OTO)(2018)

引用 1|浏览17
暂无评分
摘要
A partnership of the Gitga'at First Nation, WWF-Canada and the North Coast Cetacean Society has installed a long baseline hydrophone array in Squally Channel; a culturally, ecologically, and economically important marine environment in northern British Columbia (BC), Canada. The array consists of four synchronized bottom-mounted hydrophones that permanently record and radio-transmit data to a land-based laboratory in real-time. The array covers an area of approximately 200 km(2) to realize long-term and wide-range monitoring of marine mammals. To allow for efficient data analysis, automated detectors for cetacean vocalizations have been developed in collaboration with the University of Victoria. The detection performance has been tested using manually annotated data. We present an overview of the detectors for orca and humpback whales, and their computational and analytic performance as a function of signal-to-noise ratio. The ability to adapt to different ocean environments and target species is demonstrated by applying the detectors to 100 days of archival acoustic data recorded at two different places along the BC coast. Data recorded during pilot studies in 2017 from Squally Channel (approximately 30 days) is used to provide an overview of cetacean vocalization detections. Finally, we show acoustic activity variations and trends based on environmental factors.
更多
查看译文
关键词
marine mammal vocalizations,coastal,partnership,Gitga,First Nation,WWF-Canada,North Coast Cetacean Society,long baseline hydrophone array,Squally Channel,economically important marine environment,northern British Columbia,synchronized bottom-mounted hydrophones,radio-transmit data,land-based laboratory,wide-range monitoring,marine mammals,efficient data analysis,automated detectors,cetacean vocalizations,detection performance,manually annotated data,humpback whales,computational performance,analytic performance,signal-to-noise ratio,archival acoustic data,BC coast,cetacean vocalization detections,ocean environments,environmental factors
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要