Automated Field Photography and Cloud-Based System to Suggest Optimal Insecticide Application Timing for the Oriental Tea Tortrix, Homona magnanima Diakonoff (Lepidoptera: Tortricidae)

Abstract

We aimed to develop a chatbot application to check seasonal variations in the emergence of the oriental tea tortrix, Homona magnanima Diakonoff (Lepidoptera: Tortricidae) and the optimal timing of insecticide application for this species. The emergence was estimated via image processing. To develop an image-processing program to extract the oriental tea tortrix from the image of the sticky board, we first examined the RGB and HSV color spaces. We found that converting the HSV color space into a black and white image was effective for the model. Thereafter, we determined the optimal shooting time of the web camera by identifying the periods during which the oriental tea tortrix was not captured by the sticky board placed on the canopy surface. Although there were seasonal variations, oriental tea tortrix was not captured on the sticky board between AM4:00 and before sunrise. Thus, it can be speculated that the time period was optimal for image processing. Finally, we developed the program and linked it to a LINE app used by many tea farmers. In summary, we developed a user-friendly chatbot application that can be accessed any time without any complicated operations. This application enhances both accuracy and usability of insecticides against oriental tea tortrix infestation.