Crop load estimation in orchards: the potential of single RGB images from unmanned aerial vehicles

Summary

Throughout the entire growing season of fruit crops, from flowering, fruitlet development, ripening and harvest, to tree dormancy period, precision management is essential. It benefits agricultural production enhancement and environmental impact mitigation. The increasing popularity of unmanned aerial vehicles (UAVs) in various agricultural applications reveals the potential of UAVs in supporting orchard management, i.e., on-tree apple flower and apple counting (crop load estimation). Current solutions, yet, are generally resource-consuming and the performance is unsatisfactory. Faced with this, the main research objectives of the present thesis are: (1) identifying the current achievements and gaps of UAVs in orchard management (2) examining the possibility of crop load estimation in an apple orchard with single RGB images derived from UAVs (3) comparison of crop load estimation based on single RGB images and the estimation emanates from conventional methods.