DeepWay: A Deep Learning waypoint estimator for global path generation

Agriculture 3.0 and 4.0 have gradually introduced service robotics automation into several agricultural processes, mostly improving crops quality seasonal yield. Row-based are the perfect settings to test deploy smart machines capable of monitoring manage harvest. In this context, global path generation is essential either for ground or aerial vehicles, it starting point every type mission plan. Nevertheless, little attention has been currently given problem by research community still far be solved. order generate a viable an autonomous machine, presented proposes feature learning fully convolutional model estimating waypoints occupancy grid map. particular, we apply proposed data-driven methodology specific case row-based with general objective able cover extension crop completely. Extensive experimentation custom made synthetic dataset real satellite-derived images different scenarios proved effectiveness our demonstrated feasibility end-to-end completely planner.