基于YOLOv8的小目标检测模型的优化

doi:10.19678/j.issn.1000-3428.0070027

摘要/Abstract

摘要：

深度学习在目标检测领域的广泛应用显著提升了对大中目标的检测能力。然而, 针对小目标检测, 由于其固有的尺度小、背景复杂等挑战, 传统的目标检测算法常常会出现漏检、误检。为了提高小目标检测的精度, 对YOLOv8模型进行研究。首先, 将主干部分的卷积模块替换为RFAConv模块, 增强了模型对于复杂图像的处理能力; 其次, 在Neck部分引入混合局部通道注意力(MLCA)机制, 能够在保持计算效率的同时, 帮助模型更高效地融合不同层次的特征; 再次, 将YOLOv8的Detect头替换为Detect_FASFF头, 以解决不同特征尺度间的一致性问题, 并增强模型对小目标的检测能力; 最后, 将完全交并比(CIoU)损失函数替换为Focaler-IoU损失函数, 使模型更关注难以精确定位的小目标。实验结果显示: 改进后的模型在小目标稀疏的FloW-Img数据集上mAP@0.5提高了4.8百分点, mAP@0.5:0.95提高了3.0百分点; 在小目标密度高的VisDrone2019数据集上, mAP@0.5提升了5.9百分点, mAP@0.5:0.95提高了4.0百分点。同时还在低空数据集AU-AIR以及行人密集检测数据集WiderPerson上做了泛化对比实验。结果表明, 优化后的模型相比较原模型在小目标检测精度上有显著提升, 且适用范围更广。

关键词: 深度学习, YOLOv8网络模型, 小目标检测, 注意力机制, 损失

Abstract:

Deep learning-based object detection has significantly improved the detection of medium and large targets. However, when detecting small objects, traditional algorithms often face challenges such as missed detections and false positives owing to the inherent issues of small scale and complex backgrounds. Therefore, this study aims to enhance the accuracy of small object detection by improving the YOLOv8 model. First, the convolutional module in the backbone is replaced with the RFAConv module, which enhances the ability of the model to process complex images. Second, a Mixed Local Channel Attention (MLCA) mechanism is introduced in the neck part, allowing the model to fuse features from different layers more efficiently while maintaining computational efficiency. Third, the Detect head of YOLOv8 is replaced with the Detect_FASFF head to address the inconsistency between different feature scales and improve the ability of the model to detect small objects. Finally, the Complete Intersection over Union (CIoU) loss function is replaced with the Focaler-IoU loss function, enabling the model to focus more on small objects that are difficult to locate precisely. Experimental results show that the improved model increases mAP@0.5 by 4.8 percentage points and mAP@0.5:0.95 by 3.0 percentage points on the FloW-Img dataset, which is sparse in small objects. On the VisDrone2019 dataset which has a high density of small objects, mAP@0.5 increases by 5.9 percentage points and mAP@0.5:0.95 improves by 4.0 percentage points. In addition, generalization comparison experiments are conducted on the low-altitude dataset AU-AIR and the pedestrian-dense detection dataset WiderPerson. The optimized model significantly improves the accuracy of small object detection compared with the original model and expands its applicability.

Key words: deep learning, YOLOv8 network model, small object detection, attention mechanism, loss

王国明, 贾代旺. 基于YOLOv8的小目标检测模型的优化[J]. 计算机工程, 2025, 51(12): 294-303.

WANG Guoming, JIA Daiwang. Optimization of Small Object Detection Model Based on YOLOv8[J]. Computer Engineering, 2025, 51(12): 294-303.

https://www.ecice06.com/CN/Y2025/V51/I12/294

图/表 15

图1 YOLOv8模型

Fig.1 YOLOv8 model

图2 FASFF模块结构

Fig.2 FASFF module structure

图3 MLCA模块工作流程

Fig.3 Workflow of MLCA module

图4 改进后的网络结构

Fig.4 Structure of the improved network

图5 FloW-Img数据集上强光条件导致目标特征不明显的目标检测结果对比

Fig.5 Comparison of target detection results on the FloW-Img dataset with subtle target features caused by intense light

图6 FloW-Img数据集上遮挡场景检测结果对比

Fig.6 Comparison of target detection results on the FloW-Img dataset with obscured scene

图7 VisDrone2019数据集上夜间场景检测结果对比

Fig.7 Comparison of target detection results on the VisDrone2019 dataset with night scene

图8 VisDrone2019数据集上极小目标检测结果对比

Fig.8 Comparison of tiny objects detection results on the VisDrone2019 dataset

图9 VisDrone2019数据集上部分遮挡检测结果对比

Fig.9 Comparison of partially obscured objects detection results on the VisDrone2019 dataset

参考文献 37

1	曹家乐, 李亚利, 孙汉卿, 等. 基于深度学习的视觉目标检测技术综述. 中国图象图形学报, 2022, 27 (6): 1697- 1722.
	CAO J L , LI Y L , SUN H Q , et al. A survey on deep learning based visual object detection. Journal of Image and Graphics, 2022, 27 (6): 1697- 1722.
2	吕璐, 程虎, 朱鸿泰, 等. 基于深度学习的目标检测研究与应用综述. 电子与封装, 2022, 22 (1): 72- 80.
	LV L , CHENG H , ZHU H T , et al. Progress of research and application of object detection based on deep learning. Electronics & Packaging, 2022, 22 (1): 72- 80.
3	连哲, 殷雁君, 云飞, 等. 基于深度学习的自然场景文本检测综述. 计算机工程, 2024, 50 (3): 16- 27. doi: 10.19678/j.issn.1000-3428.0067427
	LIAN Z , YIN Y J , YUN F , et al. Review of natural scene text detection based on deep learning. Computer Engineering, 2024, 50 (3): 16- 27. doi: 10.19678/j.issn.1000-3428.0067427
4	REN S , HE K , GIRSHICK R , et al. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 39 (6): 1137- 1149.
5	HE K M, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Washington D.C., USA: IEEE Press, 2017: 2961-2969.
6	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]//Proceedings of European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 21-37.
7	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. Washington D.C., USA: IEEE Press, 2016: 779-788.
8	REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2017: 6517-6525.
9	REDMON J, FARHADI A. YOLOv3: an incremental improvement[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 89-95.
10	潘晓英, 贾凝心, 穆元震, 等. 小目标检测研究综述. 中国图象图形学报, 2023, 28 (9): 2587- 2615.
	PAN X Y , JIA N X , MU Y Z , et al. Survey of small target detection. Journal of Image and Graphics, 2023, 28 (9): 2587- 2615.
11	刘洪江, 王懋, 刘丽华, 等. 基于深度学习的小目标检测综述. 计算机工程与科学, 2021, 43 (8): 1429- 1442.
	LIU H J , WANG M , LIU L H , et al. A survey of small target detection based on deep learning. Computer Engineering and Science, 2021, 43 (8): 1429- 1442.
12	ZHANG B R , CAI H L , WEN L X . A gradient fusion-based image data augmentation method for reflective workpieces detection under small size datasets. Machine Vision and Applications, 2024, 35, 29. doi: 10.1007/s00138-024-01512-8
13	贵向泉, 刘世清, 李立, 等. 基于改进YOLOv8的景区行人检测算法. 计算机工程, 2024, 50 (7): 342- 351. doi: 10.19678/j.issn.1000-3428.0068125
	GUI X Q , LIU S Q , LI L , et al. Pedestrian detection algorithm for scenic spots based on improved YOLOv8. Computer Engineering, 2024, 50 (7): 342- 351. doi: 10.19678/j.issn.1000-3428.0068125
14	黄豪杰, 段先华, 黄欣辰. 基于深度学习水果检测的研究与改进. 计算机工程与应用, 2020, 56 (3): 127- 133.
	HUANG H J , DUAN X H , HUANG X C . Research and improvement of fruits detection based on deep learning. Computer Engineering and Applications, 2020, 56 (3): 127- 133.
15	吴双忱, 左峥嵘. 基于深度卷积神经网络的红外小目标检测. 红外与毫米波学报, 2019, 38 (3): 371- 380.
	WU S C , ZUO Z R . Small target detection in infrared images using deep convolutional neural networks. Journal of Infrared and Millimeter Waves, 2019, 38 (3): 371- 380.
16	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 7132-7141.
17	杨永波, 李栋. 改进YOLOv5的轻量级安全帽佩戴检测算法. 计算机工程与应用, 2022, 58 (9): 201- 207.
	YANG Y B , LI D . Lightweight helmet wearing detection algorithm of improved YOLOv5. Computer Engineering and Applications, 2022, 58 (9): 201- 207.
18	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of European Conference on Computer Vision. Cham, German: Springer, 2018: 3-19.
19	ZHU X K, LV S C, WANG X, et al. TPH-YOLOv5: improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2108.11539.
20	ZHANG X, LIU C, YANG D G, et al. RFAConv: innovating spatial attention and standard convolutional operation[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2304.03198.
21	WAN D H , LU R S , SHEN S Y , et al. Mixed local channel attention for object detection. Engineering Applications of Artificial Intelligence, 2023, 123, 106442. doi: 10.1016/j.engappai.2023.106442
22	ZHANG H, ZHANG S J. Focaler-IoU: more focused intersection over union loss[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2401.10525.
23	ZHENG Z H, WANG P, LIU W, et al. Distance-IoU loss: faster and better learning for bounding box regression[EB/OL]. [2024-05-22]. https://arxiv.org/abs/1911.08287.
24	LIU S T, HUANG D, WANG Y H. Learning spatial fusion for single-shot object detection[EB/OL]. [2024-05-22]. https://arxiv.org/abs/1911.09516.
25	CHENG Y, ZHU J, JIANG M, et al. Flow: a dataset and benchmark for floating waste detection in inland waters[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Washington D.C., USA: IEEE Press, 2021: 10953-10962.
26	DU D, ZHU P, WEN L, et al. VisDrone-DET2019: the vision meets drone object detection in image challenge results[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops. Washington D.C., USA: IEEE Press, 2019: 213-226.
27	BOZCAN I, KAYACAN E. AU-AIR: a multi-modal unmanned aerial vehicle dataset for low altitude traffic surveillance[C]//Proceedings of 2020 IEEE International Conference on Robotics and Automation (ICRA). Washington D.C., USA: IEEE Press, 2020: 8504-8510.
28	Zhang S , Xie Y , Wan J , et al. WiderPerson: a diverse dataset for dense pedestrian detection in the wild. IEEE Transactions on Multimedia, 2020, 22 (2): 380- 393. doi: 10.1109/TMM.2019.2929005
29	ZHANG Y F, REN W, ZHANG Z, et al. Focal and efficient IOU loss for accurate bounding box regression[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2101.08158.
30	GEVORGYAN Z. SIoU loss: more powerful learning for bounding box regression[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2205.12740.
31	FENG Z H, KITTLER J, AWAIS M, et al. Wing loss for robust facial landmark localisation with convolutional neural networks[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2018: 2235-2245.
32	REZATOFIGHI H, TSOI N, GWAK J Y, et al. Generalized intersection over union: a metric and a loss for bounding box regression[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington D.C., USA: IEEE Press, 2019: 658-666.
33	ZHANG Z . Drone-YOLO: an efficient neural network method for target detection in drone images. Drones, 2023, 7 (8): 526. doi: 10.3390/drones7080526
34	LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2209.02976.
35	WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2207.02696.
36	WANG C Y, YEH I H, LIAO H Y M. YOLOv9: learning what you want to learn using programmable gradient information[EB/OL]. [2024-05-22]. https://arxiv.org/abs/2402.13616.
37	LIANG S , WU H , ZHEN L , et al. Edge YOLO: real-time intelligent object detection system based on edge-cloud cooperation in autonomous vehicles. IEEE Transactions on Intelligent Transportation Systems, 2022, 23 (12): 25345- 25360. doi: 10.1109/TITS.2022.3158253

[1]	王舒梦, 徐慧英, 朱信忠, 黄晓, 宋杰, 李毅. 基于改进YOLOv8n的航拍轻量化小目标检测算法: PECS-YOLO[J]. 计算机工程, 2025, 51(9): 280-293.
[2]	翟志鹏, 曹阳, 沈琴琴, 施佺. 基于多时空图融合与动态注意力的交通流预测[J]. 计算机工程, 2025, 51(9): 139-148.
[3]	黄金贵, 刘朋, 唐文胜. MMD-YOLOv7:黑暗条件下车辆检测方法[J]. 计算机工程, 2025, 51(9): 340-349.
[4]	符家成, 田瑾, 张玉金, 方志军. 结合前置三元组集的知识图谱推荐[J]. 计算机工程, 2025, 51(9): 101-109.
[5]	周晨阳, 刘雪宇, 梁少华, 吴永飞. 基于Swin Transformer的肾动脉血管检测分割与定量分析[J]. 计算机工程, 2025, 51(9): 252-267.
[6]	马跃, 黄周睿, 周雯, 许艺瀚. 基于感受野注意力的轻量化林火检测算法[J]. 计算机工程, 2025, 51(9): 350-361.
[7]	徐瀅, 傅紫薇, 张伟, 陈云芳. 基于抽象语法树嵌入的智能合约漏洞检测技术[J]. 计算机工程, 2025, 51(9): 149-157.
[8]	杨明芬, 甘昀, 张兴鹏. 基于有监督自编码器的TLS加密异常流量检测[J]. 计算机工程, 2025, 51(9): 192-200.
[9]	马淦, 谷雨, 彭冬亮. 结合改进YOLOv5s和动态数据增强的海面舰船检测[J]. 计算机工程, 2025, 51(9): 294-305.
[10]	陈彦如, 刘珂良, 冉茂亮. 基于深度强化学习的外卖即时配送实时优化[J]. 计算机工程, 2025, 51(9): 328-339.
[11]	倪源松, 韩军, 邹小燕, 胡广怡, 王文帅. 两阶段自适应分块输电线路螺栓缺陷检测方法[J]. 计算机工程, 2025, 51(8): 281-291.
[12]	郝宏达, 罗健旭. 基于多尺度区域特征融合的多器官语义分割模型[J]. 计算机工程, 2025, 51(8): 270-280.
[13]	张昭理, 李家豪, 刘海, 石佛波, 何嘉文. 基于个性化遗忘建模的知识追踪方法[J]. 计算机工程, 2025, 51(8): 120-130.
[14]	闫建红, 刘芝妍, 王震. 融合时空注意力机制的多尺度卷积车辆轨迹预测[J]. 计算机工程, 2025, 51(8): 406-414.
[15]	林帆, 李建华. 基于多阶门控聚合网络的光学化学结构识别[J]. 计算机工程, 2025, 51(8): 364-372.

选择文件类型/文献管理软件名称

选择包含的内容