SCE-YOLO：改进YOLOv8的轻量级无人机视觉检测算法

doi:10.3778/j.issn.1002-8331.2411-0328

摘要/Abstract

摘要： 针对无人机航拍场景下的目标检测模型计算复杂、检测效果不佳等问题，提出一种改进YOLOv8的轻量级无人机目标检测算法SCE-YOLO。使用STA_C2f替换骨干网络中的C2f模块，提高模型的特征提取能力；将采用渐进重参数化方法改进的AIFI模块作为空间金字塔池化层，实现高质量的尺度特征交互；提出一种多尺度特征聚合扩散网络UAV_CFDPN，根据航拍小目标的尺度特征优化网络结构，设计特征聚合模块FAM以及新的特征聚合与扩散路径，使得模型获得丰富的多尺度特征和上下文信息，提高目标检测的尺度适应性；设计一种高效共享卷积模块ES-Head，在保持定位和分类能力的同时，使得模型更加轻量高效。在VisDrone2019数据集上进行测试，实验结果表明，相较于YOLOv8s，虽然提出的SCE-YOLO算法mAP50减少0.5个百分点，但参数量和计算量仅为YOLOv8s的10.0%和48.8%，在检测精度和轻量化方面相较于其他先进算法具有明显的优势。

关键词: 目标检测, YOLOv8, 多尺度特征, 特征聚合, 轻量化

Abstract: Aiming at the problems of complicated calculation and poor detection effect of target detection model in UAV aerial photography scenes, this paper proposes a lightweight UAV target detection algorithm SCE-YOLO based on improved YOLOv8. Firstly, STA_C2f is used to replace the C2f module in the backbone network to improve the feature extraction capability of the model. Secondly, the improved AIFI module using the progressive reparameterization method is used as the spatial pyramid pooling layer to achieve high-quality scale feature interaction. Thirdly, a multi-scale feature aggregation diffusion network UAV_CFDPN is proposed, which optimizes the network structure according to the scale characteristics of small aerial targets, and designs the feature aggregation module FAM and a new feature aggregation and diffusion path, so that the model can obtain rich multi-scale features and context information, and improve the scale adaptability of target detection. Finally, an efficient shared convolution module ES-Head is designed, which makes the model more lightweight and efficient while maintaining the localization and classification capabilities. Experimental results on the VisDrone2019 dataset show that compared with YOLOv8s, although the mAP50 of the SCE-YOLO algorithm proposed in this paper is reduced by 0.5 percentage points, the number of parameters and the amount of computation are only 10.0% and 48.8% of those of YOLOv8s, which has obvious advantages over other advanced algorithms in terms of detection accuracy and lightweighting.

Key words: object detection, YOLOv8, multi-scale feature, feature aggregation, lightweight

张帅, 王波涛, 涂嘉怡, 陈聪实. SCE-YOLO：改进YOLOv8的轻量级无人机视觉检测算法[J]. 计算机工程与应用, 2025, 61(13): 100-112.

ZHANG Shuai, WANG Botao, TU Jiayi, CHEN Congshi. SCE-YOLO： Improved Lightweight YOLOv8 Algorithm for UAV Visual Detection[J]. Computer Engineering and Applications, 2025, 61(13): 100-112.

参考文献

[1] XIONG X R, HE M T, LI T Y, et al. Adaptive feature fusion and improved attention mechanism-based small object detection for UAV target tracking[J]. IEEE Internet of Things Journal, 2024, 11(12): 21239-21249.
[2] ZHANG J, WAN G Y, JIANG M, et al. Small object detection in UAV image based on improved YOLOv5[J]. Systems Science & Control Engineering, 2023, 11(1): 2247082.
[3] WU H X, ZHU Y, WANG L Y. A dense small object detection algorithm based on a global normalization attention mechanism[J]. Applied Sciences, 2023, 13(21): 11760.
[4] SHEN L Y, LANG B H, SONG Z X. CA-YOLO: model optimization for remote sensing image object detection[J]. IEEE Access, 2023, 11: 64769-64781.
[5] 付锦燚, 张自嘉, 孙伟, 等. 改进YOLOv8的航拍图像小目标检测算法[J]. 计算机工程与应用, 2024, 60(6): 100-109.
FU J Y, ZHANG Z J, SUN W, et al. Improved YOLOv8 small target detection algorithm in aerial images[J]. Computer Engineering and Applications, 2024, 60(6): 100-109.
[6] WANG G, CHEN Y F, AN P, et al. UAV-YOLOv8: a small-object-detection model based on improved YOLOv8 for UAV aerial photography scenarios[J]. Sensors, 2023, 23(16): 7190.
[7] GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2015: 1440-1448.
[8] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[9] CAI Z W, VASCONCELOS N. Cascade R-CNN: delving into high quality object detection[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recog- nition. Piscataway: IEEE, 2018: 6154-6162.
[10] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proceedings of the IEEE Conference on Computer Vision. Cham: Springer, 2016: 21-37.
[11] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2999-3007.
[12] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 779-788.
[13] MU A M, WANG H J, MENG W J, et al. Small target detection in drone aerial images based on feature fusion[J]. Signal, Image and Video Processing, 2024, 18(1): 585-598.
[14] CHEN J G, WEN R H, MA L L. Small object detection model for UAV aerial image based on YOLOv7[J]. Signal, Image and Video Processing, 2024, 18(3): 2695-2707.
[15] QIU Y, FENG Y Y, CHEN H R, et al. Object detection for UAV images based on improved YOLOX[C]//Proceedings of the 2023 42nd Chinese Control Conference. Piscataway: IEEE, 2023: 7547-7553.
[16] ZHOU L M, LIU Z H, ZHAO H, et al. A multi-scale object detector based on coordinate and global information aggregation for UAV aerial images[J]. Remote Sensing, 2023, 15(14): 3468.
[17] SUN F X, HE N, LI R J, et al. GD-PAN: a multiscale fusion architecture applied to object detection in UAV aerial images[J]. Multimedia Systems, 2024, 30(3): 143.
[18] 颜豪男, 吕伏, 冯永安. 特征级自适应增强的无人机目标检测算法[J]. 计算机科学与探索, 2024, 18(6): 1566-1578.
YAN H N, LYU F, FENG Y A. Feature-level adaptive enhancement for UAV target detection algorithm[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(6): 1566-1578.
[19] LIN T Y, DOLLáR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 936-944.
[20] FENG C J, ZHONG Y J, GAO Y, et al. TOOD: task-aligned one-stage object detection[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2021: 3490-3499.
[21] MA X, DAI X Y, BAI Y, et al. Rewrite the stars[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 5694-5703.
[22] CAI X H, LAI Q X, WANG Y W, et al. Poly kernel inception network for remote sensing detection[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 27706-27716.
[23] ZHAO Y A, LV W Y, XU S L, et al. DETRs beat YOLOs on real-time object detection[C]//Proceedings of the 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2024: 16965-16974.
[24] GUO J, CHEN X, TANG Y, et al. SLAB: efficient transformers with simplified linear attention and progressive re-parameterized batch normalization[J]. arXiv:2405.11582, 2024.
[25] LIU S T, HUANG D, WANG Y H. Learning spatial fusion for single-shot object detection[J]. arXiv:1911.09516, 2019.
[26] CHEN Z X, HE Z W, LU Z M. DEA-net: single image dehazing based on detail-enhanced convolution and content-guided attention[J]. IEEE Transactions on Image Processing, 2024, 33: 1002-1015.
[27] TIAN Z, SHEN C H, CHEN H, et al. FCOS: fully convolutional one-stage object detection[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2019: 9626-9635.
[28] ZHU P F, DU D W, WEN L Y, et al. VisDrone-DET2019: the vision meets drone object detection in image challenge results[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision Workshop. Piscataway: IEEE, 2019: 213-226.
[29] 苏佳, 秦一畅, 贾泽, 等. 基于ATO-YOLO的小目标检测算法[J]. 计算机工程与应用, 2024, 60(6): 68-77.
SU J, QIN Y C, JIA Z, et al. Small object detection algorithm based on ATO-YOLO[J]. Computer Engineering and Applications, 2024, 60(6): 68-77.
[30] 孙佳宇, 徐民俊, 张俊鹏, 等. 优化改进YOLOv8无人机视角下目标检测算法[J]. 计算机工程与应用, 2025, 61(1): 109-120.
SUN J Y, XU M J, ZHANG J P, et al. Optimized and improved YOLOv8 target detection algorithm from UAV perspective[J]. Computer Engineering and Applications, 2025, 61(1): 109-120.
[31] 张光华, 李聪发, 李钢硬, 等. 基于改进YOLOv7-tiny的无人机航拍图像小目标检测算法[J/OL]. 工程科学与技术: 1-14[2024-12-02]. https://doi.org/10.15961/j.jsuese.202300593.
ZHANG G H, LI C F, LI G Y, et al. Small target detection algorithm for UAV aerial images based on improved YOLOv7-tiny[J/OL]. Advanced Engineering Sciences: 1-14[2024-12-02]. https://doi.org/10.15961/j.jsuese.202300593.
[32] OUYANG D L, HE S, ZHANG G Z, et al. Efficient multi-scale attention module with cross-spatial learning[C]//Proceedings of the 2023 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE, 2023: 1-5.
[33] SUN H, WEN Y, FENG H J, et al. Unsupervised bidirectional contrastive reconstruction and adaptive fine-grained channel attention networks for image dehazing[J]. Neural Networks, 2024, 176: 106314.
[34] WAN D H, LU R S, SHEN S Y, et al. Mixed local channel attention for object detection[J]. Engineering Applications of Artificial Intelligence, 2023, 123: 106442.
[35] XU W, WAN Y. ELA: efficient local attention for deep convolutional neural networks[J]. arXiv:2403.01123, 2024.
[36] DRAELOS R L, CARIN L. Use HiResCAM instead of Grad-CAM for faithful explanations of convolutional neural networks[J]. arXiv:2011.08891, 2021.