Intelligent Area Segmentation of Large Palm Oil Plantations with Voronoi Tessellation for Optimized UAV Flight Paths

This paper investigates the use of Unmanned Aerial Vehicles (UAVs) to automate labour-intensive tasks in commercial palm oil production. Palm oil plantations often encompass vast areas, leading to a high reliance on manual labour with industry estimates suggesting a total workforce of 500,000 workers. This high labour demand creates numerous challenges, including worker supply and strain on local resources. Consequently, Unmanned Aerial Vehicles (UAVs) have been proposed as a solution to complement some of the low-skilled tasks currently undertaken by this workforce. However, the large tracts of land typical in many of the palm oil plantations cannot be serviced by just one UAV and hence to address the limitations of U A V flight time and maximize efficiency, this research proposes a novel approach utilizing Voronoi Tessellation. This method partitions large plantations into smaller, manageable segments, each optimized for efficient UAV coverage containing an appropriate number of trees that can be comfortably serviced by the UAV model used. The effectiveness of this approach has been validated through testing on diverse datasets, demonstrating its superiority to traditional methods reliant on human expertise. Additionally, the proposed method offers the potential to minimize the total number of required segments, further enhancing efficiency. When implemented in a commercial setting, this optimized UAV deployment strategy presents significant potential for time and cost reduction, exceeding the capabilities of purely manual approaches.