RGB Image-Based Proxy Assessment of Soil Inversion, Fragmentation, and Surface Structural Roughness After Three-Body Mouldboard Ploughing: A Pilot Study

Abstract

This pilot study developed and tested a low-cost RGB image-processing workflow for assessing post-plough soil surface condition after three-body mouldboard ploughing. The study focused on deriving image-based proxy indicators of visible vegetation exposure, relative clod-size distribution, and surface structural roughness. Ten field images were captured at the College of Agricultural Engineering, Bakrajo, University of Sulaimani, Kurdistan Region, Iraq, using a Samsung Galaxy S24 Ultra smartphone camera. Visible vegetation exposure was estimated using HSV color segmentation. Soil clod features were extracted using grayscale conversion, Gaussian smoothing, Canny edge detection, and contour analysis. Surface structural roughness was estimated using an edge-density index. Results showed a mean visible vegetation exposure of approximately 1.96%, suggesting effective apparent soil inversion within the analyzed images. The mean image-derived median clod diameter was approximately 17.06 pixels, with relative clod classes dominated by fine and medium categories. The mean edge-density roughness index was approximately 0.301, indicating a highly textured post-plough surface typical of primary tillage. The proposed workflow should be interpreted as a comparative proxy-assessment method rather than a replacement for calibrated physical measurements such as sieve analysis, profilometry, LiDAR, or photogrammetric 3D reconstruction. The study demonstrates the potential of simple RGB imagery for rapid field-level tillage assessment and provides a foundation for future calibrated validation and low-cost drone-based monitoring.