ABSTRACT
Drosophila suzukii has become a serious pest in Europe attacking many soft-skinned crops such as several berry species and grapevines since its spread in 2008 to Spain and Italy. To overcome current monitoring limitations, we are developing a novel system consisting of sticky traps that are monitored by means of Unmanned Aerial Vehicles (UAVs) and an image processing pipeline that automatically identifies and counts the number of D. suzukii per trap location. To this end, we are currently collecting high-resolution RGB imagery of D. suzukii flies in sticky traps taken from both a static position (tripod) and from a UAV, which are then used as input to train deep learning models. Preliminary results show that a large part of the D. suzukii flies that are caught in the sticky traps can be correctly identified by the trained deep learning models. In the future, an autonomously flying UAV platform will be programmed to capture imagery of the sticky traps under field conditions. The collected imagery will be transferred directly to cloud-based storage for subsequent processing and analysis to identify the presence and count of D. suzukii in near real time. This data will be used as input to a decision support system (DSS) to provide valuable information for farmers.
