A Digital Orthophoto Map Generation Method Based on a Small Amount of Low Overlap Unmanned Aerial Vehicle Images

This study addresses the challenge in traditional digital orthophoto map (DOM) generation methods, which typically require a large number of highly overlapping remote sensing images, making them unsuitable for acquiring high-resolution DOM over extensive areas using multirotor unmanned aerial vehicles (UAVs) at low flight altitudes. We propose a more efficient approach that utilizes a small number of low-overlap UAV images, eliminating the need for complex structure-from-motion (SfM) processes. The method employs the Oriented FAST and Rotated BRIEF (ORB) algorithm for feature matching, leveraging its advantages in speed, rotation invariance, and robustness under low-overlap conditions. Additionally, key steps such as camera distortion correction are incorporated. Through qualitative and quantitative assessments, this method demonstrates the capability to generate high-precision orthophoto images rapidly and with minimal human intervention, thus validating its feasibility.