PEF-MODFLOW: A framework for preliminary soil profile horizon delineation based on soil color captured by smartphone images

Soil profile horizon delineation provides a vehicle for recording and communicating observations of soil profiles. A framework was developed to achieve soil profile horizon delineation based on soil color captured by smartphone images, consisting of a preprocessing module, an elbow method module, and a fuzzy c-means (FCM) algorithm module, referred to as PEF-MODFLOW. We have analyzed the effect of key parameters in PEFMODFLOW on the results and demonstrate the high performance and interpretability of the PEF-MODFLOW in the ten validation profiles. This work shows that smartphones can be an effective tool for soil profile information collection. The results of soil profile horizon delineation are still preliminary due to the limitations of the information available in the soil profile images. The flexible adjustment capability of the input data gives the PEF-MODFLOW the potential for further improvement. The PEF-MODFLOW represents a key step towards smartphone-based access to soil profile information.

Automated summary

A framework named PEF-MODFLOW based on smartphone images of soil color was developed for soil profile horizon delineation, demonstrating high performance and interpretability. The study shows smartphones can be an effective tool for soil profile information collection, while further improvements are needed for the delineation of soil profile horizons.