基于无人机遥感的河流阶地提取

河流阶地记录了河流演变过程中因构造抬升或气候变化引起的侵蚀基准面变化的信息，阶地的提取与划分对于新构造研究与古气候重建有着重要的意义，在分析传统阶地提取方法不足的基础上，提出了一种利用无人机遥感进行阶地提取的新方法.该方法利用无人机航拍影像和高精度地面控制点，通过基于计算机视觉的多视立体运动恢复结构(structure from motion with multi-view stereo, SfM-MVS)技术自动生成研究区高分辨率的正射影像和数字表面模型(digital surface model，DSM)，在此基础上根据河流阶地的几何特征，利用高程、坡度及影像灰度等统计量实现阶地信息的提取.利用这一方法对汉江支流蛮河下游河段的阶地进行了提取试验.结果表明，该方法能够提取到两级河流阶地和一级河漫滩信息，提取结果与野外实测结果有较好的一致性.与传统阶地提取方法相比，该方法具有精度好、效率高、可视化效果全面直观等优势，显示出无人机遥感在河流地貌学研究中的巨大优势和广阔应用前景. 

River Terrace Extraction Based on Unmanned Aerial Vehicle Remote Sensing

River terraces record the information of environmental changes in the process of river evolution. Proper identification and classification of terraces are of great importance to the paleo-environmental reconstruction and local tectonic evolution. Based on the analysis of drawbacks of the traditional approaches, a terrace extraction method using the unmanned aerial vehicle (UAV) remote sensing is proposed. In this paper, the UAV images and high precision ground control points are used to automatically generate the high resolution DSM (digital surface model) and orthophoto using computer vision-based structure from motion with multi-view stereo photogrammetry (SfM-MVS) techniques. On this basis, the geometric characteristics of river terraces, including the elevation, slope, and image textures are employed to delineate terraces. The proposed method is used to extract the river terraces in the lower reaches of Manhe River, a branch of Hanjiang River. The results show that the terraces extracted with the proposed method have a good agreement with those measured in the field. The proposed method has advantages of high precision and efficiency, and effective visualization for terrace representation. This method provides the technical support for the recognition of terraces, and reveals the potential of UAV remote sensing in geomorphology studies.