Computer Vision-Assisted Measurement of Ephemeral Gully Morphology Using a Portable Pin-Artboard Sensor

Soil erosion, particularly ephemeral gully (EG) erosion, poses a significant threat to agricultural sustainability and ecosystem health. Despite their substantial impact on soil degradation, EGs have been relatively understudied, primarily due to their temporary nature and the limitations of existing measurement techniques. This study introduces an integrated approach for quantifying and analyzing EGs, addressing the critical need for accurate and scalable measurement methods. Our methodology combines three key components: (1) an updated portable field tool (Gulliometer), which improves upon existing designs to enhance data collection in diverse field conditions; (2) a standardized image acquisition protocol that ensures consistent, high-quality data capture; and (3) an image processing technique leveraging easy repetitive analysis of gully cross-sections. Laboratory validation using known geometric shapes demonstrated the high precision of our methodology, with error rates below 1%. Field applications in two distinct locations in Ontario, Canada, further confirmed the practicality and effectiveness of our approach under varied environmental conditions. This approach not only advances our understanding of ephemeral gully erosion but also aids in the development of effective soil conservation strategies and informed decision-making in land management.