Object detection in aerial images is highly challenging due to the arbitrary orientation of targets and the scarcity of details in small targets. Aiming at the periodicity and boundary discontinuity of angle representation in traditional rotated bounding box detection, as well as the defect that feature representation struggles to balance spatial details and semantic enhancement, this paper proposes an Adaptive Radius Circular Smooth Label and a Lightweight Feature Enhancement Module. Detection performance is improved through geometric prior guidance and hierarchical feature interaction. Experiments on the DOTA and HRSC2016 datasets show that the proposed method achieves an mAP of 63.1% on the DOTA1.0 dataset, a 1.1% improvement over the baseline, with significantly improved detection accuracy for high-aspect ratio targets and boundary angle targets. On the HRSC2016 dataset, the mAP reaches 89.8%.