Aerial Triangulation

Aerial triangulation (AT) is a fundamental photogrammetric process aiming to determine the 3D position and orientation of images, enabling the creation of accurate 3D models from aerial imagery, primarily from UAVs. Current research focuses on improving AT efficiency and robustness, particularly through the application of deep learning architectures like convolutional neural networks (CNNs), which enhance image matching and pose estimation, often surpassing traditional methods in speed and accuracy. These advancements are significantly impacting UAV-based mapping and surveying, enabling faster and more cost-effective large-scale 3D reconstruction, and facilitating automated ground control point detection to reduce manual effort.