Abstract
The surface roughness of granular particles plays a crucial role in determining their mechanical behavior in civil engineering applications. However, the influence of image resolution on roughness characterization remains insufficiently explored. In this study, a systematic investigation is conducted to evaluate the impact of image resolution on frequency-domain and spatial-domain descriptors of particle morphology. Two-dimensional Fast Fourier Transform (FFT) analysis reveals that fine-scale texture information is progressively lost as resolution decreases. A critical threshold near 512 pixels is identified, above which the derived descriptors reliably capture intrinsic morphological features. To validate these findings in the spatial domain, a monocular 3D reconstruction pipeline is proposed and implemented using depth estimation techniques across multiple resolutions. The resulting 3D surfaces and topographic zoom-ins demonstrate that fine asperities are well preserved above the threshold but become increasingly smoothed or distorted at coarser resolutions. These results provide a quantitative workflow for selecting appropriate image resolutions in particle morphology studies and emphasize the potential risks of under-resolution in both qualitative and quantitative surface assessments.