ABSTRACT
Soil organic matter (SOM) is an important component of soil and a significant criterion in determining the dynamics of soil quality. A rapid, low-cost method to measure SOM content is needed to support the development of precision agriculture. This article studied the quantitative relationship between SOM and soil colour using a digital camera, which is relatively inexpensive and easy to operate, as a portable tool for obtaining colour information of the soil surface. The results show that mixed samples with different soil particle sizes reduce the noise of the image and are more suitable than uniform soil samples for predicting the SOM. Among the three bands of red, green, and blue (RGB), the red band had the best correlation with SOM, and its reciprocal correlation coefficient (r) reached 0.75. The reciprocal regression model of the RGB colour model provided good prediction results for mixed soil samples, with a coefficient of determination (R2) of 0.76 and a root mean square error (RMSE) of 0.55, and the validation result had an excellent predictive ability (R2val = 0.85 and RMSEval = 0.53). The single-variation predictive model of CIELa*b* colour space model through transformation of the RGB colour space model performed well. The model built by colour intensity values had a strong stability and forecasting capacity. Thus, a digital camera can be used as an alternative tool to rapidly measure SOM.
Disclosure statement
No potential conflict of interest was reported by the authors.