兰州大学机构库 >资源环境学院
基于无人机的大型食草动物调查研究-以中国青海省隆宝湿地国家自然保护区为例
Alternative TitleUnmanned aircraft system for surveying large herbivores – A case study in Longbao Wetland National Nature Reserve, Qinghai Province, China
宋清洁
Thesis Advisor周爱锋
2018-05-10
Degree Grantor兰州大学
Place of Conferral兰州
Degree Name硕士
Keyword无人机系统 大型食草动物 目视解译 种群数量估算 隆宝湿地国家自然保护区
Abstract

随着人口的不断增长和地方经济的快速发展,保护与发展的矛盾日益突出,自然保护区的生存和发展面临着严峻的考验。近年来,自然保护区内大型食草动物的数量逐渐增多,造成草地载畜压力不断加大,对草地生态系统造成严重的影响,因而急需对大型食草动物的种类和数量现状进行调查研究。但现有的大型食草动物监测技术还不足以支撑相关工作的顺利和高效开展。

本论文基于无人机系统对隆宝湿地国家自然保护区的大型食草动物进行了调查研究。通过无人机系统获取63.69平方千米的遥感影像,运用专业软件Pix4Dmapper进行了影像拼接,并结合目视解译方法识别研究区域内的大型食草动物。最后,对解译结果进行了统计分析和种群数量估算,并分析了影像拼接和无人机噪音是否对动物识别和动物行为产生影响。结果表明,无人机航拍技术非常适用于大型食草动物监测。未来无人机航拍技术将有望成为调查和监测大型食草动物数量及其分布的重要技术手段,为自然保护区管理部门更好地管理动物资源和制定相关的保护策略提供科学依据,有利于保护生物多样性、筑牢生态安全屏障、确保各类自然生态系统安全稳定以及改善生态环境质量。

通过本研究,主要得出以下结论:(1)通过目视解译方法,本论文在研究区主要发现的大型食草动物有家养藏羊和家养牦牛两种,未发现其他大型野生食草动物。由于影像分辨率较低,也未发现研究区的主要保护对象黑颈鹤。(2)统计分析表明,自然保护区的核心区和缓冲区分布有大量的家畜(家养藏羊和家养牦牛)。由此可见,自然保护区内仍有大量牧户没有迁出。按照隆宝湿地国家自然保护区的国土面积进行推算,自然保护区内大型食草动物的种群数量达23131个羊单位,且主要分布在核心区域。(3)无人机遥感影像拼接对大型食草动物的数量影响较小。对比从整体拼接影像中识别出的大型食草动物数量和从4个航带中识别出的大型食草动物数量,两者的相对误差只有1.98%;无人机系统的噪音对大型食草动物的行为几乎没有影响。与0-100米范围内识别出的大型食草动物的数量相比,100-200米的范围内识别出的大型食草动物的数量没有发生明显的变化。(4)无人机系统具有高影像分辨率、高效和可达人迹罕至的地区等优点,本研究证明无人机系统可以作为一种强有力的工具来监测大型食草动物。

Other Abstract

With the continuous growth of the population and the rapid development of the local economy, the contradiction between protection and development has become increasingly prominent. The survival and development of nature reserves are facing a severe challenge. In recent years, the population number of large herbivores in nature reserves has gradually increased.It results in increasing pressure on grassland livestock and causing serious impact on grassland ecosystems. Therefore, there is an urgent need to investigate the status and species of the large herbivores. However, the existing technologies of monitoring the large herbivores are not enough to support the smooth and efficient development of related work.

 Based on the Unmanned Aircraft System (UAS), this paper investigates the large-scale herbivores in Longbao Wetland National Nature Reserve. The remote-sensing images of 63.69 km2 were acquired by the UAS, and the images were mosaicked by the professional software Pix4Dmapper.The large herbivores in the study area were identified by visual interpretation. Finally, the statistical analysis of the interpretation results and the estimation of the population number were performed. The effects of image mosaic and the noise of UAS on animal identification were analyzed. The results show that the photography of UAS is very suitable for monitoring large herbivores. The technology for UAS is expected to become an important technical method for investigating and monitoring the population number and distribution of large herbivores and provide a scientific basis for better management of animal resources and the formulation of related protection strategies by protected area management departments, which is beneficial to the protection of biological resources in the future. It is conducive to the protection of biodiversity, the building of ecological security barriers, the safety and stability of various natural ecosystems and the improvement of the quality of the ecological environment.

Through this study, the main conclusions are as follows:(1) Through visual interpretation, the large herbivores found in the study area are Tibetan sheep and yaks in this paper. No other large wild herbivores have been found. Due to the lower resolution of the images, black-necked crane was not identified in the study area.(2) Statistical analysis shows that there are a large number of livestock (Tibetan sheep and yaks) distributed in the core area and buffer area of nature reserves. It can be seen that there are still a large number of herdsmen in the nature reserve that have not moved out. According to the national land area of the Longbao Wetland National Nature Reserve, the population number of large herbivores in the nature reserve amounts to 23131 sheep units. Figure 4-2 shows that the large herbivores in the nature reserve are mainly distributed in the core area.(3) Remote image mosaic of UAS has little effect on the population number of large herbivores. Comparing the population number of large herbivores identified from the overall mosaic image with the population number of large herbivores identified from the four flightlines, the relative error between the two was only 1.98%. The noise of UAS has little effect on the behavior of large herbivores. Compared to the population number of large herbivores identified within a range of 0-100 meter, the population number of large herbivores identified within a range of 100-200 meter did not change significantly.(4) UAS have many advantages such as high image resolution, high efficiency, and accessibility to inaccessible areas. This study demonstrates that UAS can be used as a powerful tool to monitor large herbivores.

URL查看原文
Language中文
Document Type学位论文
Identifierhttp://ir.lzu.edu.cn/handle/262010/240770
Collection资源环境学院
Recommended Citation
GB/T 7714
宋清洁. 基于无人机的大型食草动物调查研究-以中国青海省隆宝湿地国家自然保护区为例[D]. 兰州. 兰州大学,2018.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Altmetrics Score
Google Scholar
Similar articles in Google Scholar
[宋清洁]'s Articles
Baidu academic
Similar articles in Baidu academic
[宋清洁]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[宋清洁]'s Articles
Terms of Use
No data!
Social Bookmark/Share
No comment.
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.