首页|声纹识别技术支持下自然保护地鸟类多样性节律特征及监测有效性研究——以黄龙自然保护区为例

声纹识别技术支持下自然保护地鸟类多样性节律特征及监测有效性研究——以黄龙自然保护区为例

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
生物多样性是人类赖以生存和社会可持续发展的基础,是地球生命共同体的血脉和根基.声纹识别技术正在成为辅助生物多样性监测的重要手段,由于声纹数据采集的特征及计算性生物声学的发展,声学技术在物种鉴别、种群研究、环境成因研究方面正表现出优势.以黄龙自然保护区为例,以指示性物种鸟类的鸣声为对象,通过保护区中较高人为影响(黄龙景区)和较低人为影响(张家沟)这两个区域的被动式监测,采集了6 303 min声音数据.通过经典声学指数计算、人工智能物种识别与样线调查相结合的方法,研究得出:(1)人工智能技术下鸟类多样性及鸣声时间分布特征;(2)声纹识别技术较之于传统调查方法更具有效性;(3)声学指数计算对环境的响应程度与有效性辨析.研究试验性地对数据进行了分析阐释,以黄龙自然保护区为例作为方法的探索,为此类保护地保护管理提供借鉴.
Research on Bird Diversity Rhythm Features and Monitoring Effectiveness in Natural Protected Areas with Voiceprint Recognition Technology:A Case Study of Huanglong National Scenic Area
Biodiversity is the foundation for human survival and sustainable social development,embodying the lifeblood and foundation of the Earth's living community.Acoustic recognition technology is emerging as a significant tool in assisting biodiversity mon-itoring.Characterized by acoustic data collection and advanced computational bioacoustics,acoustic technology demonstrates advantages in species identification,population research,and environmental cause exploration.This study takes the Huanglong National Scenic Area as an example,collecting over 6,000 minutes of data under two distinct human impact intensities within the protected area.Through a combination of classical acoustic index calculations,artificial intelligence species identification,and line transect surveys,it is concluded that:(1)birds exhibit diverse call time distribution characteristics under artificial intel-ligence technology;(2)Acoustic recognition technology outperforms traditional survey methods in terms of effectiveness;(3)The calculation of acoustic indices demonstrates responsiveness to environmental factors and its validation.This study exper-imentally analyzes and interprets the collected data,using Huanglong as a case study to pioneer this methodological approach,making it highly valuable and applicable for conservation management in similar protected areas.

acoustic recognitionbiodiversitynature reservesoundscapeconservation management

许晓青、余楚萌、徐荣林、刘颂

展开 >

同济大学建筑与城市规划学院,自然资源部国土空间智能规划技术重点实验室,上海 200092

四川农业大学风景园林学院,成都 611130

黄龙国家级风景名胜区管理局,松潘 623306

声纹识别 鸟类多样性 自然保护地 声景 保护管理

上海市科委社会发展科技攻关计划四川省科技厅国际合作创新项目黄龙国家级风景名胜区课题自然资源部国土空间智能规划技术重点实验室项目上海市科委创新行动计划同济大学交叉科学课题(2023)

22DZ12022002022YFH0072TJW20227954012022030721DZ12030042022-3-YB-10

2024

10.12193/j.laing.2024.04.0011.002

园林
中国风景园林学会 上海市园林科学研究所

园林

影响因子：0.13
ISSN：1000-0283
年,卷(期)：2024.41(4)
  • 35