Automatic Monitoring of Pest Insects Traps Using Image Processing

Abstract

Monitoring pest insect population is currently an important issue in crop protection. At farm level insect population monitoring is routinely operated by repeated surveys by a human operator of adhesive traps, disseminated through the field, where insects remain stuck when attracted. This is a laborious and time-consuming activity, and it would be of great advantage for farmers to have an affordable system performing this task automatically. A system based on a distributed imaging device operated through a wireless network that is able to automatically acquire and transmit images of the trapping area to a remote host station is proposed. The station evaluates the insect density evolution at different farm sites and sends a message to the server that need pesticides when insect density goes over threshold or an alarming level through GSM technology. The network architecture consists of a master node hosted in a PC and a set of client nodes, spread in the fields, which act as monitoring stations. The master node coordinates the network and retrieves the captured images from the client nodes. Microcontroller based hardware system with a serial port cable and a GSM modem which is interfaced with PC is used to send messages to the particular numbers for required attention in this regard for further action.