UAV imagery has become a cost-effective tool for generating accurate 3D models and maps with various software. In this study, a comparative analysis of two widely used photogrammetry software, Agisoft Metashape and Pix4D, is conducted to evaluate their spatial accuracy in processing high-resolution UAV images. Two imaging approaches are considered: using only nadir images and incorporating both nadir and 45-degree oblique images.
The study focuses on a small 192 m² catchment in Mishor Yamin, located in the arid northeastern Negev, Israel. This region, part of the Yamin syncline, features Miocene to Neogene deposits and is subject to minimal human disturbance, making it ideal for monitoring soil erosion and surface runoff. Automated monitoring devices installed in the area, including instruments for measuring rainfall, runoff, sediment transport and dust accumulation, are used in conjunction with photogrammetry, to enhance data accuracy and support a comprehensive understanding of erosion dynamics.
Four models from each software, corresponding to two different dates and utilizing both imaging approaches, are cross-referenced. This comparison allows for a comprehensive assessment of temporal changes and spatial accuracy, providing insights into the effectiveness of each software in capturing terrain evolution and sub-centimeter-scale soil erosion.
In this study, we aim to evaluate whether accuracy is better improved through increased data by incorporating both nadir and oblique images and refining ground control point precision or by maintaining a consistent protocol across all models. This evaluation provides insights into the optimal balance between data quantity, control accuracy, and methodological consistency for effective terrain monitoring.