Dataset of dual-camera close-range photogrammetry for marker-based slope surface monitoring

This paper presents a marker-based dual-camera close-range photogrammetry dataset for field slope displacement monitoring under variable illumination conditions. The dataset contains high-resolution image sequences, extracted pixel coordinates, and synchronous Real-Time Kinematic three-dimensional coordinate ground truth acquired in a complex natural environment. The dual-camera monitoring system was used as the field data-acquisition platform, while the dataset and its associated processing products constitute the principal output of this work. The dataset is structured around two main components: (1) continuous observation data under varying illumination conditions, capturing the complete images and pixel trajectories of 14 markers over a 48-hour period that encompasses midday solar overexposure, afternoon lighting, and nighttime artificial illumination; and (2) controlled motion-experiment data, documenting the controlled movement of 13 markers across five progressive stages. For each displacement position, synchronized dual-camera images and five repeated RTK measurements were recorded. Statistical assessment of the repeated RTK measurements yielded component-wise root mean square errors of 0.435 cm, 0.398 cm, and 0.791 cm in the northing, easting, and height directions, respectively, demonstrating sub-centimeter repeatability of the reference coordinates provided in the dataset. Provided in multiple accessible formats (jpg, txt, and csv), together with Python processing scripts, this dataset can support the quantitative evaluation of photogrammetric reconstruction, marker tracking, field image registration, and environmental noise-filtering methods under realistic and variable illumination conditions.