Contactless wheat foreign material monitoring and localization with passive RFID tag arrays

In the early stage of harvesting, transportation, and storage, the grain could be mingled with clods from the field, metal pieces from aging machines, and other objects of foreign material, which will greatly influence the grain quality and food security. In this paper, we propose a novel radio frequency (RF) sensing system termed TagSee, which utilizes passive RFID tag arrays to sense the objects buried in wheat. The goal is to simultaneously sense the presence of foreign materials, locate their locations in the 3D space and sense their types, to automatically monitor the stored grain. Specifically, we use RFID received signal strength (RSS) and phase as features for foreign material detection. To design the TagSee system, we first introduce a sensing space division method. Then, an Euclidean Distance Ratio (EDR) algorithm and a heuristic method are proposed to achieve high localization success accuracy. A multi-class SVM method is used to sense different materials properties based on the impact of different foreign materials on signal features. Experimental results show that TagSee can effectively detect foreign materials in stored grain and achieve a centimeter-level localization success accuracy, and can recognize four types of objects in grain.