A Study of Determining Abnormal Behaviors by Using System for Preventing Agricultural Product Theft

Abstract

At present, the environment of agricultural production faces reduced productivity due to population decline and aging population, and agricultural products may be stolen anytime. To solve such problems, researchers have studied IoT-based smart farms. A smart farm is a system for enabling users to manage and control agricultural crops anytime anywhere on the basis of IoT (Internet of Things). This study aims to examine a system for preventing agricultural product theft by using an image monitoring system, for example, CCTV, which is part of the smart farm 1st model functions. First, the system uses image information inputted by the image monitoring system to detect objects loitering agricultural crops, and then uses directional change information to determine objects loitering the crops as abnormal behaviors. The system then uses the single-board computer, Raspberry Pi, to output real-time warning sounds and send the information of the object determined as abnormal behaviors to the user to prevent agricultural product theft.