被动声学监测设备性能比较及对鸟声识别的影响

doi:10.17520/biods.2024273

生物多样性 ›› 2024, Vol. 32 ›› Issue (10): 24273. DOI: 10.17520/biods.2024273 cstr: 32101.14.biods.2024273

被动声学监测设备性能比较及对鸟声识别的影响

黄万涛¹, 郝泽周², 张梓欣¹, 肖治术³(), 张承云¹^,^*()()

1.广州大学电子与通信工程学院, 广州 510006
2.中国林业科学研究院热带林业研究所, 广州 510520
3.中国科学院动物研究所农业虫害鼠害综合治理研究国家重点实验室, 北京 100101

收稿日期:2024-06-28 接受日期:2024-11-12 出版日期:2024-10-20 发布日期:2024-12-03
通讯作者: *E-mail: cyzhang@gzhu.edu.cn
基金资助:
国家自然科学基金(32171520)

A comparison of bird sound recognition performance among acoustic recorders

Wantao Huang¹, Zezhou Hao², Zixin Zhang¹, Zhishu Xiao³(), Chengyun Zhang¹^,^*()()

1. School of Electronics and Communication Engineering, Guangzhou University, Guangzhou 510006, China
2. Research Institute of Tropical Forestry, Chinese Academy of Forestry, Guangzhou 510520, China
3. State Key Laboratory of Integrated Management of Pest Insects and Rodents in Agriculture, Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, China

Received:2024-06-28 Accepted:2024-11-12 Online:2024-10-20 Published:2024-12-03
Contact: *E-mail: cyzhang@gzhu.edu.cn
Supported by:
National Natural Science Foundation of China(32171520)

摘要/Abstract

摘要：

被动声学监测技术能够以非侵入的方式进行长期有效的监测, 已广泛应用于鸟类的监测, 监测过程收集到的大量数据需要借助自动化识别技术进行分析处理。然而, 不同录音设备的性能差异可能会影响自动化识别软件正确识别鸟类类别的能力。本研究使用国内外6种类型录音设备对4种不同频带范围的鸟声信号进行回放录音, 选取BirdNET作为鸟类鸣声自动识别器, 对2种植被类型录音环境、5种距离和3种声源方向的回放录音信号进行鸟声识别, 评估这些变量对鸟类类别识别性能的影响。通过比较录音设备的基本参数和配置, 并构建广义线性模型(generalized linear model, GLM)对识别结果进行统计分析, 以评估不同录音设备的监测性能。结果表明,录音设备类型显著影响BirdNET对鸟类类别的识别准确率。总体上, 随着距离增加, 设备的监测有效性下降, 且在50 m或更近距离内, BirdNET的识别准确率显著更高。声源方向对识别性能也有影响, 当声源与录音设备方向相反时, 识别准确率显著下降。不同设备对4种不同频带范围鸟声信号的识别有效性存在不一致性。此外, 植被类型显著影响鸟声信号传播的衰减, 草地植被下的总体识别准确率比林地植被高40.1%。本研究建议, 在选择和部署长期录音监测设备前, 除评估成本和参数外, 还应进行实地录音监测有效性的评估。根据评估结果, 优化监测距离和方向设置, 以提升监测策略的有效性。

关键词: 被动声学监测, 声学录音设备, 鸟声识别, 深度学习, BirdNET

Abstract

Aims: Passive acoustic monitoring technology has been widely used for monitoring bird species, enabling non-invasive and long-term effective monitoring. Extensive data collection requires automated identification technologies for effective analysis. However, differences in recording device performance can affect the accuracy of automated software in identifying bird species.

Methods: Six separate recording devices from various manufacturers are tested by recording bird call playback across four frequency bands. We use BirdNET as the automatic bird sound identifier under two types of vegetation environment, five categories of distance between the recording devices and sound source, and three sound source directions. Our goal is evaluating the impact of these variables on bird species identification performance. We assess the monitoring performance of different recording devices by comparing the basic parameters and configurations of the devices and constructing a generalized linear model (GLM) to statistically analyze the identification results.

Results: Our analysis suggests the type of recording device significantly affects the ability for BirdNET to correctly identify bird species. As distance increases, the effectiveness of the devices in monitoring decreases, with the identification accuracy of BirdNET significantly higher for distances within 50 meters than beyond. Further, the direction of sound impacts identification performance, with accuracy significantly decreasing when the sound source is in opposite direction of the recording device in identifying the four types of bird sound signals with different frequency bandwidth ranges. Additionally, the vegetation type significantly affects the attenuation of bird call signals, with overall identification accuracy in grassland vegetation 40.1% higher than forest vegetation.

Conclusions: Our findings suggest the effectiveness of field recording monitoring should be assessed before selecting and deploying long-term recording monitoring equipment, in addition to evaluating equipment costs and parameters. Based on our evaluation, monitoring distance and direction settings should be optimized to enhance the effectiveness of monitoring strategies.

Key words: passive acoustic monitoring, acoustic recording devices, bird sound recognition, deep learning, BirdNET

黄万涛, 郝泽周, 张梓欣, 肖治术, 张承云 (2024) 被动声学监测设备性能比较及对鸟声识别的影响. 生物多样性, 32, 24273. DOI: 10.17520/biods.2024273.

Wantao Huang, Zezhou Hao, Zixin Zhang, Zhishu Xiao, Chengyun Zhang (2024) A comparison of bird sound recognition performance among acoustic recorders. Biodiversity Science, 32, 24273. DOI: 10.17520/biods.2024273.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.biodiversity-science.net/CN/10.17520/biods.2024273

https://www.biodiversity-science.net/CN/Y2024/V32/I10/24273

图/表 10

图1 位于广州市的两处不同植被类型的回放实验地点。(a)草地; (b)林地。

Fig. 1 Two field playback experiment sites with different vegetation types located in Guangzhou City. (a) Grassland; (b) Forest.

图2 现场回放实验整体设计示意图

Fig. 2 Schematic diagram of the overall design for field playback experiment

表1 4种鸟类鸣声信号信息

Table 1 Song signal information of four bird species

鸟类类别 Bird species	频带范围 Frequency range (kHz)	采样率 Sample rate (kHz)	位深度 Bit depth (bit)	Xeno-canto来源编号 Xeno-canto number
四声杜鹃 Cuculus micropterus	1-2	48	32	65059、290487、359666、359167、578346、822303
长尾缝叶莺 Orthotomus sutorius	2-6	48	32	835177、835194
黄胸草鹀 Ammodramus savannarum	6-11	48	32	784134、801423、825420、838342
红翅凤头鹃 Clamator coromandus	1-12	48	32	421901、421904、423390、812451

图3 4种不同鸣声频段鸟类声谱图

Fig. 3 Spectrograms of four bird species with different vocalization frequencies

图4 回放实验声源设备和录音设备现场部署情况。(a) 6种类型录音设备的录音装置; (b)声源播放装置。

Fig. 4 On-site deployment of playback experimental sound source equipment and recording devices. (a) Six types of recording equipment; (b) Sound source playback device.

表2 研究中所使用的6台录音设备的参数信息

Table 2 Specifications of the six recording devices used in the study

录音设备类型 Acoustic recorder	价格 Price (USD)	麦克风 Microphone	指向性 Directivity	声道 Audio channel	GPS	参数配置 Parameter configuration	电池类型 Battery type	频带范围 Frequency range (kHz)	采样率 Sampling rate (kHz)
Song Meter 4 (SM4)	899	外置 External	全向性 Omnidirectional	双声道 Stereo	有 Equipped	设备屏幕 Device screen	AA	0-13	8-96
Song Meter Mini (SMM)	499	外置 External	全向性 Omnidirectional	双声道 Stereo	有 Equipped	蓝牙 Bluetooth	AAA	0-11	8-96
麓音 Luyin (LY)	-	外置 External	全向性 Omnidirectional	双声道 Stereo	有 Equipped	设备屏幕 Device screen	AA	0-18	8-48
Audio Moth (AM)	97	内置 Internal	全向性 Omnidirectional	单声道 Mono	选配 Optional	蓝牙 Bluetooth	AAA	0-24	8-384
SureAnySound (SAS)	-	内置 Internal	全向性 Omnidirectional	单声道 Mono	选配 Optional	蓝牙 Bluetooth	AA	0-24	16-384
寻声 Xunsheng (XS)	-	内置 Internal	全向性 Omnidirectional	单声道 Mono	无 None	USB	钠离子电池 (18650型) Sodium-ion battery (18650 type)	0-22	44.1-384

图5 在林地环境中6种录音设备对应的鸟声识别结果(正确识别的数量)。(a)四声杜鹃; (b)红翅凤头鹃; (c)长尾缝叶莺; (d)黄胸草鹀。

Fig. 5 Bird vocalization identification results from six recording devices in forest environment. (a) Cuculus micropterus; (b) Clamator coromandus; (c) Orthotomus sutorius; (d) Ammodramus savannarum.

图6 在草地环境中6种录音设备对应的鸟声识别结果(正确识别的数量)。(a)四声杜鹃; (b)红翅凤头鹃; (c)长尾缝叶莺; (d)黄胸草鹀。

Fig. 6 Bird vocalization identification results from six recording devices in grassland environment. (a) Cuculus micropterus; (b) Clamator coromandus; (c) Orthotomus sutorius; (d) Ammodramus savannarum.

表3 广义线性模型中因变量的主效应和交互作用的结果

Table 3 Main effects and interactions of the dependent variable in the generalized linear model

变量名 Variable name	瓦尔德卡方 Wald square	自由度 Degree of freedom	P	变量名 Variable name	瓦尔德卡方 Wald square	自由度 Degree of freedom	P
录音设备类型 Acoustic recorder type	112	4	< 0.001	录音设备类型 × 距离 Acoustic recorder type × Distance	83	16	< 0.001
距离 Distance	2,275	4	< 0.001	录音设备类型 × 角度 Acoustic recorder type × Angle	20	8	< 0.05
角度 Angle	610	2	< 0.001	录音设备类型 × 鸟类类别 Acoustic recorder type × Bird species	199	12	< 0.001
植被类型 Vegetation type	1,042	1	< 0.001	录音设备类型 × 植被类型 Acoustic recorder type × Vegetation type	39	4	< 0.001
鸟类类别 Bird species	642	3	< 0.001	录音设备类型 × 距离 Acoustic recorder type × Distance	83	16	< 0.001

图7 交互作用调节分析。a、b、c、d分别为5种录音设备在距离、角度、不同频带范围发声的鸟类类别、录音环境植被类型的物种识别准确率估算平均值以及95%瓦尔德置信区间对比。横线代表估算平均值, 上下引线代表置信度区间。

Fig. 7 Interaction moderation analysis comparison of mean species identification accuracy for five recording devices across distance (a), angle (b), bird frequency bands (c), and vegetation types (d), with 95% Wald confidence intervals. Horizontal lines show mean estimates, and whiskers represent confidence intervals.

参考文献 21

[1]	Darras K, Furnas B, Fitriawan I, Mulyani Y, Tscharntke T (2018) Estimating bird detection distances in sound recordings for standardizing detection ranges and distance sampling. Methods in Ecology and Evolution, 9, 1928-1938.
[2]	Ducrettet M, Forget PM, Ulloa JS, Yguel B, Gaucher P, Princé K, Haupert S, Sueur J (2020) Monitoring canopy bird activity in disturbed landscapes with automatic recorders: A case study in the tropics. Biological Conservation, 245, 108574.
[3]	Ghani B, Denton T, Kahl S, Klinck H (2023) Global birdsong embeddings enable superior transfer learning for bioacoustic classification. Scientific Reports, 13, 22876.
[4]	Goerlitz HR (2018) Weather conditions determine attenuation and speed of sound: Environmental limitations for monitoring and analyzing bat echolocation. Ecology and Evolution, 8, 5090-5100. DOI PMID
[5]	Gupta G, Kshirsagar M, Zhong M, Gholami S, Ferres JL (2021) Comparing recurrent convolutional neural networks for large scale bird species classification. Scientific Reports, 11, 17085. DOI PMID
[6]	Hill AP, Prince P, Piña Covarrubias E, Doncaster CP, Snaddon JL, Rogers A (2018) AudioMoth: Evaluation of a smart open acoustic device for monitoring biodiversity and the environment. Methods in Ecology and Evolution, 9, 1199-1211.
[7]	Hutschenreiter A, Sosa-López JR, González-García F, Aureli F (2023) Evaluating factors affecting species detection using passive acoustic monitoring in neotropical forests: A playback experiment. Bioacoustics, 32, 660-678.
[8]	Jin YJ, Zhao LH, Qin YY, Wang JC (2023) Diversity of anurans in the Bawangling Area of Hainan National Park based on auto-recording technique. Biodiversity Science, 31, 22360. (in Chinese with English abstract) DOI
	[金彦君, 赵龙辉, 覃远玉, 汪继超 (2023) 海南国家公园霸王岭片区无尾两栖类鸣声多样性: 基于自动录音技术. 生物多样性, 31, 22360.] DOI
[9]	Kahl S, Wood CM, Eibl M, Klinck H (2021) BirdNET: A deep learning solution for avian diversity monitoring. Ecological Informatics, 61, 101236.
[10]	Liu H, Liu CH, Zhao TZ, Liu Y (2021) Bird song classification based on improved Bi-LSTM-DenseNet network. In: 4th International Conference on Robotics, Control and Automation Engineering, pp. 152-155. November 4-6, 2021, Wuhan, China.
[11]	Metcalf OC, Barlow J, Devenish C, Marsden S, Berenguer E, Lees AC (2021) Acoustic indices perform better when applied at ecologically meaningful time and frequency scales. Methods in Ecology and Evolution, 12, 421-431. DOI
[12]	Newson SE, Bas Y, Murray A, Gillings S (2017) Potential for coupling the monitoring of bush-crickets with established large-scale acoustic monitoring of bats. Methods in Ecology and Evolution, 8, 1051-1062.
[13]	Parker DM, Bernard RTF (2018) The use of acoustic detectors for assessing bat species richness and functional activity in a South African national park. Mammalia, 83, 53-63.
[14]	Pérez-Granados C (2023a) BirdNET: Applications, performance, pitfalls and future opportunities. Ibis, 165, 1068-1075.
[15]	Pérez-Granados C (2023b) A first assessment of birdnet performance at varying distances: A playback experiment. Ardeola, 70, 257-269.
[16]	Pérez-Granados C, Bota G, Giralt D, Albarracín J, Traba J (2019) Cost-effectiveness assessment of five audio recording systems for wildlife monitoring: Differences between recording distances and singing direction. Ardeola, 66, 311.
[17]	Prior L, Aland K, Levengood AL, Potvin DA (2023) Vocal activity of the Eastern Ground Parrot (Pezoporus wallicus wallicus) and implications for acoustic monitoring efforts. Emu-Austral Ornithology, 123, 364-369.
[18]	Rhinehart TA, Chronister LM, Devlin T, Kitzes J (2020) Acoustic localization of terrestrial wildlife: Current practices and future opportunities. Ecology and Evolution, 10, 6794-6818. DOI PMID
[19]	Winiarska D, Szymański P, Osiejuk TS (2024) Detection ranges of forest bird vocalisations: Guidelines for passive acoustic monitoring. Scientific Reports, 14, 894.
[20]	Xu ZH, Liu SY, Zhao Y, Tu WQ, Chang ZF, Zhang ET, Guo J, Zheng D, Geng J, Gu GY, Guo CP, Guo LL, Wang J, Xu CY, Peng C, Yang T, Cui MQ, Sun WC, Zhang JT, Liu HT, Ba CQ, Wang HQ, Jia JC, Wu JZ, Xiao C, Ma KP (2020) Evaluation of the identification ability of eight commonly used plant identification application softwares in China. Biodiversity Science, 28, 524-533. (in Chinese with English abstract) DOI
	[许展慧, 刘诗尧, 赵莹, 涂文琴, 常诏峰, 张恩涛, 郭靖, 郑迪, 耿鋆, 顾高营, 郭淳鹏, 郭璐璐, 王静, 徐春阳, 彭钏, 杨腾, 崔梦琪, 孙伟成, 张剑坛, 刘皓天, 巴超群, 王鹤琪, 贾竞超, 武金洲, 肖翠, 马克平 (2020) 国内8款常用植物识别软件的识别能力评价. 生物多样性, 28, 524-533.] DOI
[21]	Zhong EZ, Guan ZH, Zhou XC, Zhao YJ, Li H, Tan SB, Hu KR (2021) Application of passive acoustic monitoring technology in the monitoring of western black crested gibbons. Biodiversity Science, 29, 109-117. (in Chinese with English abstract)
	[钟恩主, 管振华, 周兴策, 赵友杰, 李函, 谭绍斌, 胡坤融 (2021) 被动声学监测技术在西黑冠长臂猿监测中的应用. 生物多样性, 29, 109-117.]

被动声学监测设备性能比较及对鸟声识别的影响

A comparison of bird sound recognition performance among acoustic recorders

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 21

相关文章 15

编辑推荐

Metrics

本文评价

[1]	白皓天, 余上, 潘新园, 凌嘉乐, 吴娟, 谢恺琪, 刘阳, 陈学业. AI辅助识别的鸟类被动声学监测在城市湿地公园中的应用[J]. 生物多样性, 2024, 32(8): 24188-.
[2]	王永财, 万华伟, 高吉喜, 胡卓玮, 孙晨曦, 吕娜, 张志如. 基于深度学习的我国北方常见天然草地植物识别[J]. 生物多样性, 2024, 32(4): 23435-.
[3]	郭倩茸, 段淑斐, 谢捷, 董雪燕, 肖治术. 鸟声标注技术及其在被动声学监测中的应用[J]. 生物多样性, 2024, 32(10): 24313-.
[4]	陈蕾, 许志勇, 苏菩坤, 赖小甜, 赵兆. 依频声学多样性指数用于人类活动区域的适用能力[J]. 生物多样性, 2024, 32(10): 24286-.
[5]	刘莹莹, 龚立新, 曾皓, 冯江, 董永军, 王磊, 江廷磊. 被动声学监测在蝙蝠研究中的应用[J]. 生物多样性, 2024, 32(10): 24233-.
[6]	申小虎, 李冠宇, 史洪飞, 王传之. 数据不平衡下鸟声识别的集成学习策略[J]. 生物多样性, 2024, 32(10): 24215-.
[7]	李乐, 张承云, 裴男才, 高丙涛, 王娜, 李嘉睿, 武瑞琛, 郝泽周. 基于被动声学监测技术的城市绿地景观格局与鸟类多样性关联分析[J]. 生物多样性, 2024, 32(10): 24296-.
[8]	郝泽周, 张承云, 李乐, 高丙涛, 曾伟, 王淳, 王梓炫, 黄万涛, 张悦, 裴男才, 肖治术. 城市鸟类多样性被动声学监测与评价技术应用[J]. 生物多样性, 2024, 32(10): 24123-.
[9]	谢将剑, 沈忱, 张飞宇, 肖治术. 融合音频及生态位信息的跨地域鸟类物种识别方法[J]. 生物多样性, 2024, 32(10): 24259-.
[10]	蔡建民, 何培宇, 杨智鹏, 李露莹, 赵启军, 潘帆. 基于深度特征融合的鸟鸣识别方法及其可解释性分析[J]. 生物多样性, 2023, 31(7): 23087-.
[11]	黄雨菲, 路春燕, 贾明明, 王自立, 苏越, 苏艳琳. 基于无人机影像与面向对象-深度学习的滨海湿地植物物种分类[J]. 生物多样性, 2023, 31(3): 22411-.
[12]	申小虎, 朱翔宇, 史洪飞, 王传之. 基于机器学习鸟声识别算法研究进展[J]. 生物多样性, 2023, 31(11): 23272-.
[13]	肖治术, 崔建国, 王代平, 王志陶, 罗金红, 谢捷. 现代生物声学的学科发展趋势及中国机遇[J]. 生物多样性, 2023, 31(1): 22423-.
[14]	马海港, 范鹏来. 被动声学监测技术在陆生哺乳动物研究中的应用、进展和展望[J]. 生物多样性, 2023, 31(1): 22374-.
[15]	吴科毅, 阮文达, 周棣锋, 陈庆春, 张承云, 潘新园, 余上, 刘阳, 肖荣波. 基于音节聚类分析的被动声学监测技术及其在鸟类监测中的应用[J]. 生物多样性, 2023, 31(1): 22370-.