Pests detection system for agricultural crops using intelligent image analysis

The paper proposes an intelligent method for real-time monitoring of the evolution of pests in an agricultural crop. The system consists of a pheromone trap, a camera that periodically fetches frames with trap and a data transceiver module where the frame is encapsulated with the date and station id. The data arrives at a server where they are retrieved and analyzed using two deep learning artificial neural networks. One determines the type and number of pests and the other identifies the evolution of the pest population. The system has been experimented with an apple tree culture in order to early identify pests and combat them.