1. 中科北纬 (北京) 科技有限公司, 北京 100043
2. 江西省林学会, 江西南昌 330046
| 摘 要: | 本研究旨在探讨人工智能技术在湿地鸟类监测中的应用,以提高监测效率和准确性。研究结合现代通信网络、人工智能等高新技术,构建覆盖关键生态系统的天空地一体化监测网络。通过在湖北沉湖湿地部署高清摄像头、声音传感器和无人机等设备,实现对鸟类活动的实时视频监控和声音数据采集。利用深度学习算法对视频流和声音信号进行分析,自动识别鸟类的种类和行为模式。研究选择了11种鸟类作为监测对象,通过无人机采集了大量影像数据,并使用YOLOv8和PP-HGNET算法进行鸟类识别和分类。结果表明,AI监测系统在鸟类识别精度上达到了87.7%的均值平均精度,能够为鸟类动态监测、疫源疫病监测、保护研究等提供可靠的数据支撑。此外,AI技术还能监测湿地环境指标,为生态保护提供科学依据。研究证明,AI技术在湿地鸟类监测中具有广阔的应用前景,能够推动监测技术的革新,提高生态保护的效率和效果。 |
| 关 键 词: | 人工智能; 湿地鸟类; 智能监测 |
| DOI: | 10.57237/j.jaf.2025.01.002 |
1. China Natural-harmony Blueprint Technology Co., Ltd, Beijing 100043, China
2. Jiangxi Society of Forestry, Nanchang 330046, China
| Abstract: | This study aims to explore the application of Artificial Intelligence (AI) technology in wetland bird monitoring to improve monitoring efficiency and accuracy. By integrating modern communication networks and AI technologies, a sky-ground integrated monitoring network covering key ecosystems has been established. In the Chen Lake Wetland in Hubei, high-definition cameras, sound sensors, and drones were deployed to achieve real-time video surveillance and sound data collection of bird activities. Deep learning algorithms were utilized to analyze video streams and sound signals, automatically identifying bird species and behavior patterns. The study selected 11 bird species as monitoring objects and collected a large amount of image data using drones. YOLOv8 and PP-HGNET algorithms were employed for bird recognition and classification. The results showed that the AI monitoring system achieved a mean Average Precision (mAP) of 87.7% in bird recognition accuracy, providing reliable data support for bird dynamic monitoring, epidemic source monitoring, and conservation research. Additionally, AI technology can monitor wetland environmental indicators, offering a scientific basis for ecological protection. The research demonstrates that AI technology has broad application prospects in wetland bird monitoring, capable of driving the innovation of monitoring technologies and enhancing the efficiency and effectiveness of ecological conservation. |
| Keywords: | Artificial Intelligence; Wetland Birds; Intelligent Monitoring |
| 1. | 湖北省野生动物疫源疫病智慧监测项目 (20230810-000539). |
| [1] | Tian, S., J. Xu, J. Li, Z. Zhang, and Y. Wang, Research advances of Galliformes since 1990 and future prospects. Avian research, 2018. 9(1): p. 1-13. |
| [2] | 姚毅与彭祥林. 洞庭湖水鸟监测与研究. in 湖泊保护与生态文明建设——中国湖泊论坛. 2014. |
| [3] | 颜凤等, 围填海对湿地水鸟种群、行为和栖息地的影响. 生态学杂志, 2017. 36(7): 第7页. |
| [4] | 蒋敏等, 鸟类全景观测系统中的人工智能识别技术. 浙江林业科技, 2021. |
| [5] | 提浩, 自然场景下的人脸检测及表情识别算法研究, 2018, 北京交通大学. |
| [6] | 王虎诚等, 被动声学监测技术在九里湖湿地公园鸟类监测中的应用研究. 江苏林业科技, 2023. 50(3): 第30-36页. |
| [7] | 王庆合等, AI鸟类监测识别系统在内乡湍河湿地的应用分析. 河南林业科技, 2024. 44(01): 第39-40+43页. |
| [8] | 汪洋, 基于深度学习的细粒度鸟类识别方法研究与系统实现, 2020, 南昌大学. |
| [9] | 周铁林, 基于深度学习的北方湿地鸟类识别方法研究, 2021, 沈阳理工大学. |
| [10] | 张雅千与韦璐璐, 人工智能理念在热带景观小区中的应用——以海南三亚凤凰山居为例. 世界生态学, 2022. 11(4): 第7页. |
| [11] | 汤静雯, 赖惠成与王同官, 远距离情形下的改进YOLOv8行人检测算法. 计算机工程, 2024. |
| [12] | 杨硕等, 基于路测图像与改进ResNet50网络的自动驾驶场景天气识别算法. 汽车与新动力, 2024. 7(02): 第15-22页. |
| [13] | 晏宏亮, 基于深度学习的东洞庭湖鸟类识别研究, 2022, 湖南农业大学. |
| [14] | 王元卓, 靳小龙与程学旗, 网络大数据:现状与展望. 计算机学报, 2013. 36(6): 第1125-1138页. |
| [15] | 吕丹阳等, 生成式人工智能在公共服务中应用的机遇与挑战. 电子科技大学学报社科版, 2024. 26(3): 第1-11页. |