Towards automatic monitoring of insect pests using IoT camera-equipped pheromone traps: a case study for Spodoptera litura (Lepidoptera: Noctuidae)

Abstract

To efficiently monitor insect pests, we developed an electronic device that incorporates a pheromone trap and an Internet of things (IoT) camera (E-trap) that can semi-automatically check the daily number of captured insects. Images captured by the IoT camera were transmitted over a 4G wireless network, and the number of insects captured by the trap was visually confirmed from the images. Our E-trap and two types of conventional (funnel-type and dry-type) traps were placed in two soybean fields in Japan in 2021. The trends of Spodoptera litura (Fabricius) (Lepidoptera: Noctuidae) caught daily by the E-trap and those observed at five-day intervals by the conventional traps were compared. Although the general trends in the occurrence of S. litura were similar to the data obtained among the different trapping methods, the data obtained daily by the E-trap showed a more detailed change in trap catches. The coefficient of determination between the number of S. litura visually estimated from the images and the actual number of S. litura caught was greater than 0.95. This demonstrates the practicality of the proposed E-trap, which minimizes the need for monitoring and maintaining the insect traps.