Non-Invasive Computer Vision-Based Fruit Fly Larvae Differentiation: Ceratitis capitata and Bactrocera zonata

The Mediterranean fruit fly (Ceratitis capitata) and the peach fruit fly (Bactrocera zonata) are two of the most economically significant agricultural pests affecting fruit production worldwide. Both are considered quarantine pests in several countries, which oblige the use of restrictive measures to assure safe trade with countries where these flies are present. As the quarantine status of these two pests is not similar in every country, discriminating measures among these two fruit flies' larvae in the exported fruits is critical for safe trade. Traditional DNA-based detection methods, though accurate, are costly and time-consuming, while manual morphological identification is practically impossible. In this study, we propose a novel non-invasive method utilising computer vision for rapid differentiation between larvae of these two species based on a short video recording of a single larva freely moving on a Petri dish. Our results reveal good separation between the two species with 90% accuracy using videos as short as 15 s long.