• 首页关于本刊期刊订阅编委会作者指南过刊浏览
苏晓雪,李希来,孙华方,李成一,宋梓涵.2023.青藏高原高原鼠兔活动节律监测方法研究.动物学杂志,58(4):630-639.
青藏高原高原鼠兔活动节律监测方法研究
Study on the Monitoring Method of Plateau Pika Activity Rhythm on the Qinghai-Tibet Plateau
投稿时间:2022-12-14  
DOI:10.13859/j.cjz.202304016
中文关键词:  高原鼠兔  高寒草甸  监测系统  活动节律
英文关键词:Plateau Pika  Alpine meadow  Monitoring system  Activity rhythm
基金项目:青海大学青年教师科研基金项目(No. 2021-QNY-12),国家自然科学联合基金项目(No. U21A20191),三江源国家公园联合研究专项(No. LHZX-2020-08),高等学校学科创新引智计划项目(No. D18013),青海省科技创新创业团队项目“三江源生态演变与管理创新团队”;
作者单位
苏晓雪 青海大学农牧学院 西宁 810016 
李希来 青海大学农牧学院 西宁 810016 
孙华方 青海大学农牧学院 西宁 810016 
李成一 青海大学农牧学院 西宁 810016 
宋梓涵 青海大学农牧学院 西宁 810016 
摘要点击次数: 54
全文下载次数: 551
中文摘要:
      高原鼠兔(Ochotona curzoniae)是高寒草甸生态系统特有物种和关键物种,其挖掘觅食等活动行为直接影响本地高寒草甸的健康状况和持续发展。以有效洞口调查为代表的传统人工观察方法很难连续数天观察,造成可获取数据量小、效率较低,导致高原鼠兔行为生物节律研究数据采集困难。因此本文构建了融合图像自动采集、优化数字图像处理算法和智能识别技术的高原鼠兔监测系统,可无接触、数字化、连续性监测高原鼠兔行为活动。首先,考虑高原鼠兔活动节律监测需求,在青海省河南蒙古族自治县克旗河滩布局并划分了面积为25 m × 30 m的实验小区;其次,监测实验区域,获取高原鼠兔全周期活动节律数据;最后提出基于改进帧差法的高原鼠兔目标自动识别策略。实验结果显示,本文设计的监测系统可在自然环境下完成高原鼠兔活动数据采集,基于改进帧差法的高原鼠兔目标自动识别策略精准率、召回率、F1值分别为87.20%、83.37%、0.85,可准确获取高原鼠兔详细的活动节律信息。活动节律分析结果表明,高原鼠兔地面活动高峰期在10:00时和16:00时,夜晚无活动,属于白昼型动物。青藏高原鼠兔活动节律监测方法可有效提升高原鼠兔的种群密度、活动节律等相关研究效率和准确度。
英文摘要:
      [Objectives] Plateau Pika (Ochotona curzoniae) is an endemic and key species in the alpine meadow ecosystem, and its behaviors of digging, foraging, and moving directly affect the health and stability of the alpine meadow. However, it is difficult to achieve continuous behavioral observation for several days to obtain all the data required for the biological rhythm observation of Plateau Pika. Due to the shortcoming of the traditional method,a Plateau Pika monitoring system is constructed by integrating image automatic acquisition, optimizing digital image and processing algorithm, and intelligent recognition technology, which realize non-contact, continuous, and digital monitoring of Plateau Pika. [Methods] According to the need for Plateau Pika activity rhythm monitoring, the experimental layout was designed with an area of 25 m × 30 m in Henan Mongolian Autonomous County, Qinghai Province. The Plateau Pika monitoring system is designed, and the optimized and improved frame difference method is proposed. [Results] Through the monitoring of the experimental area, the video image data of the recorded Plateau Pika is obtained. The Plateau Pika monitoring system meets the basic application requirements of observing Plateau Pika activities in the natural environment. The improved frame difference method is used to realize the detection of moving targets based on the complex image. The average accuracy rate, recall rate, and weighted average were 87.20%, 83.37%, and 0.85 (Table 2). The peak of ground activity of Plateau Pika appeared at 10:00 and 16:00, and it was inactive at night (Fig. 6), so it belonged to diurnal animals. [Conclusion] These results can meet the target detection requirements of Plateau Pika’s monitoring system, recording the activity rhythm and frequency, which provides new research ideas and effective technical means for Plateau Pika behavior research.
附件
查看全文  查看/发表评论  下载PDF阅读器