藏区高原典型环境地形目标的轻量化检测模型

doi:10.3785/j.issn.1008-973X.2026.03.015

浙江大学学报(工学版)

2026, Vol. 60

Issue (3): 594-603 DOI: 10.3785/j.issn.1008-973X.2026.03.015

计算机技术、控制工程

藏区高原典型环境地形目标的轻量化检测模型

雒伟群1(

),陆敬蔚1,吴佳缔1,梁钰迎1,申传鹏1,朱睿1,2,*(

)

1. 西藏民族大学信息工程学院，陕西咸阳 712082
2. 厦门大学航空航天学院，福建厦门 361102

Lightweight detection model for typical environmental terrain target in Tibetan Plateau

Weiqun LUO1(

),Jingwei LU1,Jiadi WU1,Yuying LIANG1,Chuanpeng SHEN1,Rui ZHU1,2,*(

)

1. College of Information Engineering, Xizang Minzu University, Xianyang 712082, China
2. School of Aerospace Engineering, Xiamen University, Xiamen 361102, China

全文: PDF(3893 KB) HTML

摘要：

为了解决通用航空飞行器在高原山区应用场景中，所面临的传统方法障碍物识别精度低和计算量过大等问题，提出障碍物检测模型AO-YOLO. 通过优化现有的YOLOv8n模型，引入多尺度扩张注意力机制，增强Neck网络对不同尺度特征的融合能力. 采用HGBlock结构替换原有的Bottleneck模块，通过轻量级卷积建立特征层次化关系，使网络能够同时提取局部与全局上下文信息. 结合内存高效的注意力模块与轻量级卷积模块，设计新的检测头，降低模型参数量与计算成本. 实验结果显示，在构建的高原山区障碍物数据集（FOD）上，相较于YOLOv8n，AO-YOLO的mAP@0.5指标提升了2.7%，mAP@0.5:0.95指标提升了2.0%，整体计算量减少了24.7%. 该模型在高原航空障碍物检测任务中具有精度较高和轻量化的特性.

关键词： 高原航空; 障碍物检测; 轻量化YOLOv8n; 多尺度扩张注意力; 特征提取

Abstract:

An obstacle detection model AO-YOLO was proposed in order to solve the problems of low accuracy and excessive calculation of obstacle recognition in traditional methods faced by general aviation aircraft in plateau mountain application scenarios. A multi-scale expanded attention mechanism was introduced by optimizing the existing YOLOv8n model in order to enhance the ability of Neck network to fuse features of different scales. HGBlock structure was used to replace the original Bottleneck module, and lightweight convolution was used to establish the feature hierarchical relationship, so that the network can extract local and global context information at the same time. The memory-efficient attention module and lightweight convolution module were combined to design a new detection head in in order to reduce the number of model parameters and computational cost. The experimental results showed that the mAP@0.5 index of AO-YOLO increased by 2.7% and the mAP@0.5:0.95 index increased by 2.0% on the constructed plateau mountain obstacle dataset (FOD) compared with YOLOv8n, and the overall calculation amount reduced by 24.7%. The proposed model has the characteristics of high accuracy and lightweight in the highland aviation obstacle detection task.

Key words: plateau aviation obstacle detection lightweight YOLOv8n multi-scale dilated attention feature extraction

收稿日期: 2025-01-30 出版日期: 2026-02-04

TP 183

基金资助: 西藏自治区重点研发计划资助项目（XZ202401ZY0102,XZ202403ZY0019,XZ202402ZY0017）；厦门市自然科学基金资助项目（3502Z20227179）；教育部人文社会科学规划基金资助项目（23XZJAZH001）；西藏自治区自然科学基金重点资助项目（XZ202401ZR0055）；福建省自然科学基金资助项目（2022J01058）；中国航空科学基金资助项目（2023Z032068001）；水声对抗技术重点实验室基金资助项目（JCKY2024207CH05）；西藏民族大学基金资助项目（Y2024050）.

通讯作者: 朱睿 E-mail: 1034228464@qq.com;zhurui@xmu.edu.cn

作者简介: 雒伟群（1969—），男，教授，硕导，从事机器学习、大数据与知识工程的研究. orcid.org/0009-0007-7608-3358. E-mail：1034228464@qq.com

	服务
	把本文推荐给朋友
	加入引用管理器
	E-mail Alert
	作者相关文章
	雒伟群
	陆敬蔚
	吴佳缔
	梁钰迎
	申传鹏
	朱睿

引用本文:

雒伟群,陆敬蔚,吴佳缔,梁钰迎,申传鹏,朱睿. 藏区高原典型环境地形目标的轻量化检测模型[J]. 浙江大学学报(工学版), 2026, 60(3): 594-603.

Weiqun LUO,Jingwei LU,Jiadi WU,Yuying LIANG,Chuanpeng SHEN,Rui ZHU. Lightweight detection model for typical environmental terrain target in Tibetan Plateau. Journal of ZheJiang University (Engineering Science), 2026, 60(3): 594-603.

链接本文:

https://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2026.03.015 或 https://www.zjujournals.com/eng/CN/Y2026/V60/I3/594

图 1 AO-YOLO网络的结构图

图 2 C2f_MS模块的结构

图 3 扩张率控制感受野的示意图

图 4 PP-HGNetV2结构图

图 5 HGBlock模块的结构图

图 6 重构后的骨干网络结构

图 7 SADetect的结构

表 1 训练参数的设置

表 2 在FOD数据集上进行的模型对比实验

图 8 YOLOv8n的精确度-召回率曲线

图 9 AO-YOLO的精确度-召回率曲线

表 3 AO-YOLO模型的消融实验

图 10 不同注意力机制的对比

表 4 在VisDrone2019上的实验结果

表 5 在NWPU VHR-10上的实验结果

图 11 检测结果的可视化对比

1	DALAL N, TRIGGS B. Histograms of oriented gradients for human detection [C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego: IEEE, 2005: 886–893.
2	GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus: IEEE, 2014: 580–587.
3	GIRSHICK R. Fast R-CNN [C]//Proceedings of the IEEE International Conference on Computer Vision. Santiago: IEEE, 2016: 1440–1448.
4	REN S, HE K, 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 doi: 10.1109/TPAMI.2016.2577031
5	HE K, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN [C]//Proceedings of the IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2980–2988.
6	赵志宏, 郝子晔改进YOLOv8的航拍小目标检测方法: CRP-YOLO[J]. 计算机工程与应用, 2024, 60 (13): 209- 218 ZHAO Zhihong, HAO Ziye Improved YOLOv8 aerial small target detection method: CRP-YOLO[J]. Computer Engineering and Applications, 2024, 60 (13): 209- 218
7	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector [C]// 14th European Conference on Computer Vision. Cham: Springer, 2016: 21–37.
8	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 779–788.
9	TANG S, ZHANG S, FANG Y. HIC-YOLOv5: improved YOLOv5 for small object detection [C]//Proceedings of the IEEE International Conference on Robotics and Automation. Yokohama: IEEE, 2024: 6614-6619.
10	YANG Y. Drone-view object detection based on the improved YOLOv5 [C]//Proceedings of the IEEE International Conference on Electrical Engineering, Big Data and Algorithms. Changchun: IEEE, 2022: 612–617.
11	韩俊, 袁小平, 王准, 等基于YOLOv5s的无人机密集小目标检测算法[J]. 浙江大学学报: 工学版, 2023, 57 (6): 1224- 1233 HAN Jun, YUAN Xiaoping, WANG Zhun, et al UAV dense small target detection algorithm based on YOLOv5s[J]. Journal of Zhejiang University: Engineering Science, 2023, 57 (6): 1224- 1233
12	张徐, 朱正为, 郭玉英, 等基于cosSTR-YOLOv7的多尺度遥感小目标检测[J]. 电光与控制, 2024, 31 (4): 28- 34 ZHANG Xu, ZHU Zhengwei, GUO Yuying, et al Multi-scale remote sensing small target detection based on cosSTR-YOLOv7[J]. Electronics Optics and Control, 2024, 31 (4): 28- 34
13	王燕妮, 张婧菲. 改进YOLOv8的无人机小目标检测算法 [J/OL]. 探测与控制学报, 2024, 46(6): 1–10. (2024-12-12). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=XDYX20241211005&dbname=CJFD&dbcode=CJFQ. WANG Yanni, ZHANG Jingfei. A modified YOLOv8 algorithm for UAV small object detection [J/OL]. Journal of Detection and Control, 2024, 46(6): 1–10. (2024-12-12). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=XDYX20241211005&dbname=CJFD&dbcode=CJFQ.
14	WANG G, CHEN Y, 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 doi: 10.3390/s23167190
15	李岩超, 史卫亚, 冯灿面向无人机航拍小目标检测的轻量级YOLOv8检测算法[J]. 计算机工程与应用, 2024, 60 (17): 167- 178 LI Yanchao, SHI Weiya, FENG Can Lightweight YOLOv8 detection algorithm for small object detection in UAV aerial photography[J]. Computer Engineering and Applications, 2024, 60 (17): 167- 178
16	龙伍丹, 彭博, 胡节, 等基于加强特征提取的道路病害检测算法[J]. 计算机应用, 2024, 44 (7): 2264- 2270 LONG Wudan, PENG Bo, HU Jie, et al Road damage detection algorithm based on enhanced feature extraction[J]. Journal of Computer Applications, 2024, 44 (7): 2264- 2270
17	JIAO J, TANG Y M, LIN K Y, et al DilateFormer: multi-scale dilated transformer for visual recognition[J]. IEEE Transactions on Multimedia, 2023, 25: 8906- 8919 doi: 10.1109/TMM.2023.3243616
18	徐佩, 陈亚江融合Swin Transformer的YOLOv5口罩检测算法[J]. 智能计算机与应用, 2024, 14 (5): 83- 92 XU Pei, CHEN Yajiang Mask detection algorithm based on YOLOv5 integrating Swin Transformer[J]. Intelligent Computer and Applications, 2024, 14 (5): 83- 92
19	李登峰, 高明, 叶文韬结合轻量级特征提取网络的舰船目标检测算法[J]. 计算机工程与应用, 2023, 59 (23): 211- 218 LI Dengfeng, GAO Ming, YE Wentao Ship target detection algorithm combined with lightweight feature extraction network[J]. Computer Engineering and Applications, 2023, 59 (23): 211- 218
20	YU H, WAN C, LIU M, et al. Real-time image segmentation via hybrid convolutional-transformer architecture search [EB/OL]. (2024-03-15). https://arxiv.org/abs/2403.10413.
21	徐光达, 毛国君多层级特征融合的无人机航拍图像目标检测[J]. 计算机科学与探索, 2023, 17 (3): 635- 645 XU Guangda, MAO Guojun Aerial image object detection of UAV based on multi-level feature fusion[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17 (3): 635- 645
22	张洋, 夏英多尺度特征融合的遥感图像目标检测方法[J]. 计算机科学, 2024, 51 (3): 165- 173 ZHANG Yang, XIA Ying Object detection method with multi-scale feature fusion for remote sensing images[J]. Computer Science, 2024, 51 (3): 165- 173
23	曹利, 徐慧英, 谢刚, 等. ASOD-YOLO: 基于YOLOv8n改进的航空小目标检测算法 [J/OL]. 计算机工程与科学, 2024, 46(9): 1–13. (2024-09-27). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=JSJK20240906002&dbname=CJFD&dbcode=CJFQ. CAO Li, XU Huiying, XIE Gang, et al. ASOD-YOLO: improved aviation small objection detection algorithm based on YOLOv8n [J/OL]. Computer Engineering and Science, 2024, 46(9): 1–13. (2024-09-27). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=JSJK20240906002&dbname=CJFD&dbcode=CJFQ.
24	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection [C]//Proceedings of the IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2999–3007.
25	CAI Z, VASCONCELOS N. Cascade R-CNN: delving into high quality object detection [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 6154–6162.
26	LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications [EB/OL]. (2022-09-07). https://arxiv.org/abs/2209.02976
27	PAN X, GE C, LU R, et al. On the integration of self-attention and convolution [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 805–815.
28	YANG L, ZHANG R Y, LI L, et al. SimAM: a simple, parameter-free attention module for convolutional neural networks [C]//Proceedings of the International Conference on Machine Learning. Graz: ACM, 2021: 11863-11874.

[1]	周著国,鲁玉军,吕利叶. 基于改进YOLOv5s的印刷电路板缺陷检测算法[J]. 浙江大学学报(工学版), 2025, 59(8): 1608-1616.
[2]	付家瑞,李兆飞,周豪,黄惟. 基于Convnextv2与纹理边缘引导的伪装目标检测[J]. 浙江大学学报(工学版), 2025, 59(8): 1718-1726.
[3]	梁礼明,龙鹏威,金家新,李仁杰,曾璐. 基于改进YOLOv8s的钢材表面缺陷检测算法[J]. 浙江大学学报(工学版), 2025, 59(3): 512-522.
[4]	傅幼萍,张航,厉梦菡,孟濬. 基于脉搏波信号多维度特征的身份识别[J]. 浙江大学学报(工学版), 2025, 59(3): 566-576.
[5]	王博特,王卿,刘强,金波. 基于多通道振动主元特征的风电机组叶片自监督异常识别方法[J]. 浙江大学学报(工学版), 2025, 59(3): 653-660.
[6]	陈江浩,杨军. 结合深度可分离卷积的多源遥感融合影像目标检测[J]. 浙江大学学报(工学版), 2025, 59(12): 2545-2555.
[7]	杨硕,娄旭,刘硕,李佳睿,王磊. 基于非参数脑网络的帕金森亚型临床差异分析[J]. 浙江大学学报(工学版), 2025, 59(11): 2451-2458.
[8]	林俊杰,朱雅光,刘春潮,刘昊洋. 面向移动作业的腿足机器人数字孪生系统[J]. 浙江大学学报(工学版), 2024, 58(9): 1956-1969.
[9]	王海军,王涛,俞慈君. 基于递归量化分析的CFRP超声检测缺陷识别方法[J]. 浙江大学学报(工学版), 2024, 58(8): 1604-1617.
[10]	韩康,战洪飞,余军合,王瑞. 基于空洞卷积和增强型多尺度特征自适应融合的滚动轴承故障诊断[J]. 浙江大学学报(工学版), 2024, 58(6): 1285-1295.
[11]	钟博,王鹏飞,王乙乔,王晓玲. 基于深度学习的EEG数据分析技术综述[J]. 浙江大学学报(工学版), 2024, 58(5): 879-890.
[12]	罗钒睿,刘振宇,任佳辉,李笑宇,程阳. 基于改进卡尔曼滤波的轻量级激光惯性里程计[J]. 浙江大学学报(工学版), 2024, 58(11): 2280-2289.
[13]	蒋林,刘林锐,周安娜,韩璐,李平原. 基于运动预测的改进ORB-SLAM算法[J]. 浙江大学学报(工学版), 2023, 57(1): 170-177.
[14]	卞艳,宫雨生,马国鹏,王昶. 基于无人机遥感影像的水体提取方法[J]. 浙江大学学报(工学版), 2022, 56(4): 764-774.
[15]	徐泽鑫,段立娟,王文健,恩擎. 基于上下文特征融合的代码漏洞检测方法[J]. 浙江大学学报(工学版), 2022, 56(11): 2260-2270.

Viewed

Full text

Abstract

Cited

Shared

Discussed