An Efficient Object Detection Algorithm for UAV-Borne Platforms

Efficient object detection on unmanned aerial vehicle (UAV) platforms is a challenging task due to factors such as small object scales, complex environmental backgrounds, and the resource constraints of UAV-Borne platforms. To address these challenges, this paper introduces the SPD-Conv and SNLB modules into the detection network, significantly enhancing the model’s robustness and detection accuracy. By implementing layer fusion and mixed precision quantization methods, we maintain model accuracy while improving inference speed after deployment. Experimental results demonstrate that our algorithm, when deployed on the RK3588 platform, improves detection accuracy by 2.8% and 1.9% compared to EdgeYOLO and FFCA-YOLO, respectively, while also showing superior real-time performance and energy efficiency.