Detection and Counting of Fruit Flies (Diptera: Tephritidae) in a Smart Trap

Economically important fruit flies cause billions of dollars in direct losses in the world's fruit industry and limit the commercialization of fruit. Monitoring these pests is an important phytosanitary measure, which helps in efficient decision-making regarding management. Thus, our objective was to develop an automatic counting system, through video tracking, embedded in a McPhail trap, to assist in the automatic monitoring of fruit flies. The programming of the automatic fruit fly counting system in the trap was developed in Python language, using the Background Subtraction method available in the OpenCV library. With regard to the recognition efficiency of the counting system, we report scores of 90%. This level of accuracy of the system is ideal to take the trap to the field in future investigations. This technology is very promising for the early detection of the level of pest control, as it is monitored remotely and in real-time, avoiding frequent visits to the field.