引用本文:纪景纯,赵 原,邹晓娟,宣可凡,王伟鹏,刘建立,李晓鹏.无人机遥感在农田信息监测中的应用进展[J].土壤学报,2019,56(4):773-784. DOI:10.11766/trxb201811190508
JI Jingchun,ZHAO Yuan,ZOU Xiaojuan,XUAN Kefan,WANG Weipeng,LIU Jianli,LI Xiaopeng.Adcancement in Application of UAV Remote Sensing to Monitoring of Farmlands[J].Acta Pedologica Sinica,2019,56(4):773-784. DOI:10.11766/trxb201811190508
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 501次   下载 1274 本文二维码信息
码上扫一扫!
分享到: 微信 更多
无人机遥感在农田信息监测中的应用进展
纪景纯,赵 原,邹晓娟,宣可凡,王伟鹏,刘建立,李晓鹏
1.中国科学院南京土壤研究所;2.山西大学环境与资源学院
摘要:
快速实时地掌握农田信息是实施精准农作的基础。以无人机为平台的低空遥感探测技术,具有空间分辨率高、时效性强和成本低等特点,可填补地面监测和高空遥感间的测量尺度空缺,因此在农田信息精准监测领域具有广泛的应用前景。近年来,随着无人机飞行平台稳定性增强、操作难度降低,机载遥感设备的轻量化和多样化,以及遥感数据处理技术的进步,无人机遥感在农田信息监测领域得到了快速发展。本文对国内外相关研究成果进行了总结,对常用遥感技术类型和数据处理方法以及具体应用方向和实施效果进行了综述,并提出了当前存在的突出问题和未来的发展方向,以期为推动无人机遥感在农田信息监测和精准农业中更广泛的应用提供依据。
关键词:  低空遥感  无人机  农田信息监测  作物估产  生长诊断
基金项目:国家重点研发计划项目(2016YFD0300601),中国科学院南京土壤研究所“一三五”计划和领域前沿项目(ISSASIP1661)
Adcancement in Application of UAV Remote Sensing to Monitoring of Farmlands
JI Jingchun1, ZHAO Yuan2, ZOU Xiaojuan1, XUAN Kefan3, WANG Weipeng2, LIU Jianli1, LI Xiaopeng1
1.Institute od soil science,Chinese academy of science;2.College of Environmental & Resource Science of Shanxi University;3.Institute of Soil Science, Chinese Academy of Sciences
Abstract:
Fast and real-time acquisition of farmland information is the basis of precision farming. The technology of using unmanned aerial vehicle (UAV) as a platform for low-altitude remote sensing of farmlands features high spatial resolution, real-time and low cost, and can be used to fill the gas between field survey and high-altitude remote sensing in measuring scale. Therefore, the technology has a wide prospect in application to accurate monitoring of farmlands for real-time information. In recent years, with the technology advancing rapidly, UAV is more stable in flight and easier to operate, and airborne remote sensing equipment is getting light and diversified, capable of acquiring different remote sensing information such as visible light data, multispectral information, hyperspectral data and three-dimensional point cloud data. Moreover, recent development of the remote sensing data processing technology has enabled investigators to process data faster and more accurate. Hyperspectral data is well known to be abundant in information and hard to process, too. Fortunately, a number of dimensionality reduction methods have been developed. Comentropy is the most straightforward method. Principal component analysis and independent component analysis are also widely used to choose mean wave bands or vegetation indexes. There is a wavelet basis more practical for wavelet decomposition method too. Thermal infrared data is a powerful tool to reflect field temperature and moisture, and just because of this, it can also objectively reflect crop diseases and pests. Because the monitoring can cover an area as large as an entire tract of fields, even the conventional multispectral data can be interpreted into accurate information of crop growth. Three-dimensional point cloud is a newly booming data in farmland monitoring and capable of providing information of plant height and accurate position of the plant as well. So more information is available for assessing crop growing conditions. Three-dimensional point cloud can be obtained by LiDAR. The first step to process this kind of data is to remove and smooth noise points caused by system error or overlight of target surface or accidental factors, fix vulnerable and missing parts of the point cloud, and in the end fine field model is achieved. SfM features three-dimensional point cloud with real color and can be used to build a field model with visible-light images. This method uses two images as initial homologous photographs to establish a three-dimensional coordinate by feature matching, front and back rendezvous calculation. Then add in new images, find new matchable feature points and adjust them till optimum, Repeat the steps until all images are added in. All these extraordinary technological advances together make the application of UAV-based remote sensing possible in precision farming. Unmanned flying technology reduces monitoring costs for it needs less fuel and no pilot. And investigators can conduct monitoring more frequently to enhance timeliness. Suitable flight height combined with high-performance sensor makes it possible to acquire data superior in spatial resolution. For the features given above UAV-based remote sensing can fill up the gap between ground monitoring and high-altitude remote sensing, such as satellite remote sensing, in the measuring scale. Therefore, the technology has a bright application prospect in application to gathering information for precision farming. In addition this paper has also summarized findings and achievements in related researches at home and abroad, introduced the commonly used remote sensing technologies and data processing methods, as well as specific application directions and their implement effects are discussed too. This paper also addresses the existing problems and future development directions, in an attempt to promote the usage of UAV remote sensing in farmland monitoring. It is hoped that this technology can be more widely used in precision agriculture.
Key words:  Low altitude remote sensing  Unmanned aerial vehicle (UAV)  Farmland information monitoring  Yield estimation  Growth diagnosis