The Development of a new virtual croplands erosion measurement system using three-dimensional laser scanner and empirical Kostiakov-Lewis models

Abstract The traditional method for the estimation of soil loss by croplands erosion is the measurement through erosion pins, which consists of stakes with a specified height distributed across the land area; however, this method is inaccurate to measure the soil depth. In some cases, a simple description of the 3D and depth map scenarios is useful to understand the problem. In this study I introduced an instrument of three-dimensional measurement with a camera and a laser to extract the information from the soil depth. Then, a structural mesh in 3D from the scene is obtained and this information is processed in order to estimate the soil loss. The retrieved depth depends on the dynamic relationships among the objects in the scene and the instrument control; this device can adapt to the graphics and dimensions of the surface under study. Here I show the generation of a depth map from a mesh with third order polynomials. This mesh describes the behavior of the surface where full panoramic integration is performed by using multiple partial views of object. The results show that this system has the capacity to acquire soil erosion information; in addition, other parameters can be obtained such as visual information of texture, size perspective, and shadows edges. In this work the vision system covered an area of 1.20 m long by 0.8 m wide. This laser device offered depth analysis and topology from the soil using computer vision through empirical Kostiakov-Lewis models.