Novel measurements and features for the characterization of soil surface roughness

Agricultural soils provide a variety of ecological services, including nutrient cycling, water purification and storage, carbon sequestration, and flood protection. Soil Surface Roughness (SSR) represents a key parameter for evaluating the terrain quality structure, especially in the layers beneath the usual primary tillage depth, and therefore its estimation has been widely investigated over the years in many fields of science and engineering. This paper proposes the adoption of an innovative sensing approach that relies on contactless measurements provided by a depth camera in contrast to traditional contact measurements such as pin meter and roller chain. In addition, novel features are investigated to achieve a complete statistical description of the SSR with the least number of parameters. The proposed methods are validated in an experimental campaign performed on a vineyard plot. This research could be useful for many applications, including soil erosion prediction models, autonomous vehicle navigation in rural and agricultural settings, and controlled traffic farming.