Predicting soil organic matter from cellular phone images under varying soil moisture

Measuring soil organic matter is costly, difficult, time-consuming and labour-intensive using traditional soil sampling and laboratory analyses. Organic matter may be measured visually using image processing technology and machine learning. Soil color could be accurately measured with cell phone images at soil moisture contents less than 10%. Cellular phone images showed strong potential as proximal soil sensor.