| 摘要: |
| 【目的】利用小型消费级无人机航拍获取地物影像,通过地物阴影、高度差、色差快速提取地物,进而获取地物结构信息。【方法】本研究选择乌蒙山区会泽县为研究区域,针对冬闲田闲置土地资源、种植结构相对单一的区域展开试验,利用高分辨率无人机遥感影像对燕麦进行识别,同时结合超声波传感器数据估算地物高度,并与实际高度和无人机生成的传统测高方法得到的高度进行相关性分析,获取高精度、可靠性强的数据。【结果】试验结果表明基于可见光燕麦的总体分类精度为91.46%,Kappa系数为0.857,在增加DSM数据后的分类总体精度为98.91%,Kappa系数为0.982,燕麦生长模型的模拟精度为92.8%。研究表明由无人机获取的代表作物冠层高度信息的DSM数据能够显著提升燕麦的识别效果,依赖于光谱和高程信息的识别精度更高,识别结果也更可靠。上述结果表明该方法能够高精度计算地物高度,并且其精度超过无人机传统测高方法生成数字表面模型提取的地物高度。【结论】该研究提出的小型消费级无人机利用地物阴影计算高度方法能够应用于实际,并使得未布设控制点的无人机数据也可作为阴影测高法的数据来源,为遥感快速获取地物高度信息提供了一种新的思路。 |
| 关键词: 燕麦 无人机 高度 快速 高精度 |
| DOI： |
| 分类号: |
| 基金项目:云南省“科技入滇”项目“乌蒙山区燕麦提质增效与产品研发关键技术研究与示范”(202003AD150016)、云南省“徐丽君专家工作站”经费(202005AF150074)、现代农业产业技术体系建设专项资金(Cars-34)资助。第一 |
|
| PHANTOM 4 RTK+ DJI image processing technology in oat growth simulation application |
|
Xu Lijun1, Zhang Deqi2, Xue Wei3, Nie Yingying1, Rao Xiong4, Yang Jiaxia1, Xu Shuhua4, Zhu Meng4, Fu Yanfei4, Qiao Zhenglin4
|
|
1.Institute of Agricultural Resources and Regional Planning,Chinese Academy of Agricultural Sciences/Hulunbuir Grassland Ecosystem Observation And Research Station,CAAS Beijing;2.Shenyang Tianjun Houde Communication Network Engineering Co,LTD,Shenyang Liaoning;3.Qingdao Agricultural University;4.Yunnan Province Qujing Huize County Agriculture Bureau
|
| Abstract: |
| [Purpose] Using small consumer UAV aerial photography to acquire ground object image, quickly extract ground object through ground object shadow, height difference, and color difference, and then obtain ground object structure information.[Method] In this study, Huize County in the Wumeng Mountain area was selected as the research area, and the experiment was carried out in the area where the idle land resources of winter fallow field and the planting structure was relatively single. The high-resolution UAV remote sensing image was used to identify oat, and the height of the ground object was estimated combined with ultrasonic sensor data. The correlation analysis with the actual height and the height obtained by the traditional altimetry method generated by UAV is carried out to obtain high-precision and reliable data.[Result] The experimental results showed that the overall classification accuracy of oat-based on visible light was 91.46%, and the Kappa coefficient was 0.857. After adding DSM data, the overall classification accuracy was 98.91%, the Kappa coefficient was 0.982, and the simulation accuracy of the oat growth model was 92.8%. The study showed that DSM data, which represents crop canopy height information obtained by UAV, could significantly improve the identification effect of oat, and the identification accuracy depending on spectral and elevation information was higher, and the identification results were more reliable. The above results show that this method can calculate the height of a ground object with high precision, and its accuracy exceeds that of a ground object height extracted by a digital surface model generated by the UAV traditional altimetry method.[Conclusion] The proposed method for calculating the height of ground object shadow by small consumer UAV can be applied in practice, and the UAV data without control points can also be used as the data source of shadow height measurement method, which provides a new idea for obtaining the height information of ground object quickly by remote sensing. |
| Key words: oats UAV height fast high precision |