A smartphone-based soil color sensor: For soil type classification.

Soil type is a key indicator in field survey, but the current soil classification method largely depends on personal experiences of operators. Because different soils have different soil colors, soil can be classified by measuring soil colors. In this paper, we compared and analyzed the roles of the visible spectrum and machine vision adopted in soil classification and proposed a new smartphone-based, low-cost, and miniaturized soil color classification sensor. The CMOS device of the mobile phone was directly used as the sensor and the flashgun of the phone was used as the light source. The peripheral components included external lens, shading devices, and color calibration card, which were assembled on the phone directly. The colors of soil and proofread cards were acquired by the smartphone and converted into RGB signals. With RGB signals, after simple processing, rapid soil classification could be achieved. (C) 2016 Elsevier B.V. All rights reserved.