Path Planning for Quadrotor's Collision Avoidance Using a Single Field Programmable Gate Array Platform
Tin-Yu Chou, Cheng-Min Jen, Ying-Hao Yu · 2025
A collision avoidance system is critical to safe navigation for unmanned aerial vehicles (UAVs), especially in complicated terrain like farmland. Existing solutions often integrate multiple computational platforms, increasing payload and power consumption. This paper proposes a novel, lightweight, real-time path-planning system based on a single fieldprogrammable gate array (FPGA) platform and a monocular camera for onboard processing. A key innovation of this work is introducing the Rotation Matrix and Arc Path Planning (RMAPP) algorithm, which optimizes computational efficiency and ensures precise obstacle avoidance. Compared to conventional methods, RMAPP significantly reduces processing overhead by abridging complex trigonometric calculations, making it more suitable for resource-constrained embedded platforms. Theoretical justifications and comparative data demonstrate that this method enhances quadrotor maneuverability while reducing power consumption. Experimental results validate the efficiency of the proposed approach in real-world farmland environments. This research will contribute to advanced quadrotor navigation in the future.