INTELLIGENT AERIAL SEEDING SYSTEM USING DEEP LEARNING AND ADAPTIVE VISUAL TRACKING

Forestation in Sudan is hindered by desertification, harsh climate, and the remoteness of arid regions, leading to low seedling survival and inefficient resource use. This study presents an Intelligent Aerial Seeding System that integrates Unmanned Aerial Vehicles (UAVs) with computer vision and machine learning to enhance the precision and effectiveness of aerial seeding. Real-time vision techniques, including object tracking, are employed to monitor seed trajectories during descent and provide feedback for in-flight corrections using machine learning-based controllers. A convolutional neural network (CNN) is used to classify soil types from aerial imagery, enabling dynamic adjustment of seeding force according to local terrain. Numerical simulations in MATLAB and Simulink model seed trajectories under varying crosswind conditions. Tests with biodegradable mud-coated Senegalia senegal seeds show improved aerodynamic stability, soil penetration, and placement accuracy. The proposed UAV-based system reduces seed loss, increases operational efficiency, and enables scalable, data driven reforestation, supporting the United Nations Sustainable Development Goals on climate action (SDG 13) and life on land (SDG 15).