Estimating soil moisture content using red-green-blue imagery from digital camera

Abstract

Soil moisture content is one of the most important soil physical characteristics and very beneficial for plant growth. Instead of employing laboratory analysis, estimating soil moisture in more simple and practical method, such as using a digital camera is very important to improve the efficiency of agricultural productivity. Air and soil temperature, which strongly relate to soil moisture are indicating the climate anomaly. Therefore, it is necessary to observe soil moisture using red green image from camera. This research aimed at estimating soil moisture from several soil types using Red-Green-Blue (RGB) imagery from digital camera. Eight soil samples from several locations of Alfisol soil types were collected, air-dried and sieved. Similar volume of water was applied at each soil sample, then all samples arranged for picture capture. Soon after the picture was captured using RGB digital camera, the soil moisture was investigated using gravimetric method. The image of the captured soil samples then was processed using ENVI software for interpretation. Correlation and regression analysis were applied to obtain the estimation model and accuracy level. The results found that RGB image taken from digital camera can be used to estimate soil moisture with moderate accuracy.