The accuracy of soil moisture prediction using an RGB camera on maize and peanut plantation

Abstract

Forest observation to use soil moisture meters for monitoring soil moisture are still relatively inefficient if done on very large land because it still relies on relatively much energy and requires a long time. This study aims to determine the accuracy of the use of aerial photographs from the Red-Green-Blue (RGB) camera Drone to estimate soil moisture levels. Aerial photographs were analyzed with the ImageJ application to find digital numbers, and actual soil moisture measurements by the gypsum block method. Soil moisture and digital numbers are then analyzed by regression and correlation. The results show that this method has a low accuracy. However, the comparison between the actual soil moisture and the soil moisture from the prediction doesn’t show a significant difference. The accuracy of the estimation depends on the camera settings and the weather, so further calibration and testing are necessary if used under different conditions.