UAV Photogrammetry with Low Image Overlap for Panoramic Imaging and River Morphology Monitoring

Abstract

River morphology plays a crucial role in our understanding of geomorphological processes and the transmission of geographical information. High-resolution images captured by unmanned aerial vehicles (UAV) have emerged as indispensable tools for monitoring river morphologies, providing detailed insights into the structure and dynamics of river systems. However, traditional image mosaic algorithms, which heavily rely on feature point matching, encounter significant challenges when dealing with highly similar scenes, such as those found in river environments. To address this limitation, this study introduces a robust and innovative method for computing panoramic images independently of any feature point matching. The proposed method consists of several key steps. Firstly, a UAV flight strategy was carefully designed to facilitate the acquisition of low-overlap aerial images, ensuring efficient and effective data collection. Secondly, a panoramic image was computed through point matching based on a plane model reconstructed by image projection. In this step, a generalized lighting ray model was developed to accurately facilitate the 3D-to-2D transformation, ensuring the precision of the panoramic image. Finally, the computed panoramic image was utilized to evaluate river morphology, providing valuable insights into the structure and dynamics of the river system. In experimental fields, UAV images with 10% overlap were collected and processed using the proposed algorithm. Experimental results demonstrate that the proposed algorithm can generate high-resolution panoramic images with remarkable clarity and detail. Additionally, the target river information was successfully extracted from the panoramic images, paving the way for further morphological analysis and research.