Quantifying Soil Particulate Organic Matter Using Image Analysis
The particulate fraction of soil organic matter (POM) is of interest due to its sensitivity to management and contributions to soil structure and biological integrity. This work explored an imaging-based alternative to laboratory-based estimates of POM that are laborious and time-consuming. This novel approach only requires lightly prepared samples to be saturated and photographed in a table-top imaging box. Samples were illuminated with Light Emitting Diodes (LED) spanning limited sections of visible to near-infrared (Vis-NIR) spectrum. Linear regression showed estimates of POM area, determined using image analysis, were linearly related to POM concentrations in sieved soils. The predictive accuracy of image-based estimates increased with POM fraction size and varied by residue type.
