UAV-based warehouse inventory object detection algorithm using improved YOLOv8

doi:10.3785/j.issn.1008-973X.2026.12.010

Journal of ZheJiang University (Engineering Science)

2026, Vol. 60

Issue (12): 2642-2651 DOI: 10.3785/j.issn.1008-973X.2026.12.010

UAV-based warehouse inventory object detection algorithm using improved YOLOv8

Xiaoping YUAN(

),Xinwen ZHENG,Cong PENG

School of Information and Control Engineering, China University of Mining and Technology, Jiangsu 221000, China

Download:

HTML

PDF(2161KB) HTML
Export: BibTeX | EndNote (RIS)

Abstract

An improved object detection algorithm was proposed aiming at the issue of insufficient recognition accuracy of the original YOLOv8n algorithm in the task of drone-based warehouse inventory. Inner-EIoU, which was adaptable to target overlap and relaxed aspect ratio constraints, was introduced as the bounding box loss function in order to enhance the recognition capability for densely stacked goods. The standard convolution in the C2f module was replaced with omni-dimensional dynamic convolution that incorporated a multi-dimensional attention mechanism in the feature extraction stage. The FocalModulation method was applied to the spatial pyramid pooling-fast module in order to improve feature representation. A channel pruning method based on the LAMP criterion was used for model lightweighting in order to reduce the parameter scale of the model and increase inference speed. The experimental results on a self-built warehouse dataset showed that the improved algorithm achieved mAP50 of 94.2%, with model size of only 2.8 MB and inference speed of 135 frames/s, demonstrating the feasibility for deployment on edge computing device.

Key words： drone object detection warehouse inventory attention mechanism lightweight model

Received: 30 September 2025 Published: 21 October 2026

CLC:

TP 391

Fund: 国家级大学生创新训练项目（202410290039Z）.

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Xiaoping YUAN
	Xinwen ZHENG
	Cong PENG

Cite this article:

Xiaoping YUAN,Xinwen ZHENG,Cong PENG. UAV-based warehouse inventory object detection algorithm using improved YOLOv8. Journal of ZheJiang University (Engineering Science), 2026, 60(12): 2642-2651.

URL:

https://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2026.12.010 OR https://www.zjujournals.com/eng/Y2026/V60/I12/2642

基于改进YOLOv8的无人机仓储盘点检测算法

针对原始YOLOv8n算法在无人机仓储货物盘点任务中识别精度不足的问题，提出改进的目标检测算法. 为了提升对密集堆叠货物的识别能力，引入能够适应目标重叠情况并放宽目标框宽高比约束的Inner-EIoU作为边界框损失函数. 在特征提取部分，将C2f模块中的标准卷积替换为结合多维注意力机制的全维度动态卷积，在快速空间金字塔池化模块中采用FocalModulation方法，以增强特征表达能力. 为了减少模型参数规模并提高推理速度，使用基于LAMP剪枝准则的通道剪枝方法对模型进行轻量化处理. 在自建仓储数据集上的实验结果表明，改进后算法的mAP50达到94.2%，模型大小仅为2.8 MB，推理速度提升至135帧/s，验证了该方法在边缘计算设备上部署的可行性.

关键词： 无人机, 目标检测, 仓储盘点, 注意力机制, 轻量化模型

Fig.1 Network structure of YOLOv8

Fig.2 Comparison of structure between traditional convolution and omni-dimensional dynamic convolution

Fig.3 Activation mode of attention mechanism in four dimensions

Fig.4 Structure of FocalModulation network

Fig.5 Schematic diagram of bounding box, anchor box and internal area in Inner-IoU object detection

Fig.6 Flowchart of LAMP pruning

Tab.1 Dataset category and bounding box information

Tab.2 Experimental result of loss function

Tab.3 Experimental result of attention module

Tab.4 Experimental result of SPPF module

Tab.5 Ablation experiment result of YOLOv8n

Fig.7 Training curve of mAP50 for original YOLOv8n model and improved model

Tab.6 Performance comparison of LAMP pruning algorithm under different speed-up

Fig.8 Comparison of mAP50 of different algorithms

Tab.7 Performance comparison of different algorithms

Fig.9 Comparison of detection performance between improved and original models in various scenarios


[1]	刘飞, 孙延明, 陈娟, 等新时代工业4.0背景下加快制造业向智能制造转型的策略研究: 以粤港澳大湾区为例[J]. 广东经济, 2019, (3): 72- 81 Liu Fei, Sun Yanming, Chen Juan, et al Strategic research on accelerating the transformation of manufacturing industry to intelligent manufacturing under the background of new era industry 4.0: taking the Guangdong-Hong Kong-Macao greater bay as an example[J]. Guangdong Economy, 2019, (3): 72- 81 doi: 10.3969/j.issn.1005-0759.2019.03.014

[2]	何凡. 智能仓储管理系统的设计与实现 [D]. 西安: 西安电子科技大学, 2019: 23–24. He Fan. Design and implementation of an intelligent warehouse management system [D]. Xi’an: Xi'an University of Electronic Science and Technology, 2019: 23–24.

[3]	孙其博, 刘杰, 黎羴, 等物联网: 概念、架构与关键技术研究综述[J]. 北京邮电大学学报, 2010, 33 (3): 1- 9 Sun Qibo, Liu Jie, Li Shan, et al Internet of things: summarize on concepts, architecture and key technology problem[J]. Journal of Beijing University of Posts and Telecommunications, 2010, 33 (3): 1- 9 doi: 10.3969/j.issn.1007-5321.2010.03.001

[4]	王文娟基于无线射频识别技术的数字化物流仓储管理系统[J]. 科学技术与工程, 2019, 19 (2): 170- 174 Wang Wenjuan Digital logistics warehouse management system based on radio frequency identification technology[J]. Science Technology and Engineering, 2019, 19 (2): 170- 174 doi: 10.3969/j.issn.1671-1815.2019.02.029

[5]	袁瑞阳, 刘新宇, 刘贤栋, 等. 无人云库存盘点系统[C]//物联网与无线通信-2018年全国物联网技术与应用大会论文集. 杭州: 人民邮电出版社, 2018. Yuan Ruiyang, Liu Xinyu, Liu Xiandong, et al. Unmanned cloud library storage point system [C]// Internet of Things and Wireless Communication-Proceedings of the 2018 National Internet of Things Technology and Application Conference. Hangzhou: Posts and Telecom Press, 2018.

[6]	王杰. 基于机器视觉的四旋翼无人机仓库盘点系统研究与实现 [D]. 绵阳: 西南科技大学, 2021: 45−55. Wang Jie. Research and implementation of a four rotor unmanned aerial vehicle warehouse inventory system based on machine vision [D]. Mianyang: Southwest University of Science and Technology, 2021: 45−55.

[7]	赵滨淋, 陈功, 李胜基于自适应特征融合的无人机小目标检测算法[J]. 山东农业大学学报（自然科学版）, 2024, 55 (6): 814- 825 Zhao Binlin, Chen Gong, Li Sheng Uav small target detection algorithm based on adaptive feature fusion[J]. Journal of Shandong Agricultural University (Natural Science Edition), 2024, 55 (6): 814- 825

[8]	刘智慧, 张泉灵大数据技术研究综述[J]. 浙江大学学报（工学版）, 2014, 48 (6): 957- 972 Liu Zhihui, Zhang Quanling Research overview of big data technology[J]. Journal of Zhejiang University (Engineering Science), 2014, 48 (6): 957- 972

[9]	张慧, 赵东辉基于低空经济的智能物流系统运输路径优化[J]. 物流科技, 2025, 48 (21): 84- 88 Zhang Hui, Zhao Donghui Optimization of intelligent logistics system transportation path based on low altitude economy[J]. Logistics Sci-tech, 2025, 48 (21): 84- 88

[10]	DHL“抢跑”无人机送货 [J]. 英才, 2014(12): 16. DHL “rush to run” drone delivery [J]. Talents Magazine, 2014(12): 16.

[11]	Howard A G, Zhu M, Chen B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications [PP/OL]. arXiv (2017-04-17)[2025-09-15]. https://arxiv.org/abs/1704.04861.

[12]	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.

[13]	Li C, Zhou A, Yao A. Omni-dimensional dynamic convolution [C]// 2022 International Conference on Learning Representations Conference, 2022: 3-5.

[14]	Yang J, Li C, Dai X, et al. Focal modulation networks [PP/OL]. arXiv (2022-03-22)[2025-09-15]. https://arxiv.org/abs/2203.11926.

[15]	Zhang H, Xu C, Zhang S. Inner-IoU: more effective intersection over union loss with auxiliary bounding box [PP/OL]. arXiv (2023-11-06)[2025-09-15]. https://arxiv.org/abs/2311.02877.

[16]	Lee J, Park S, Mo S, et al. Layer-adaptive sparsity for the magnitude-based pruning [PP/OL]. arXiv (2020-10-15)[2025-09-15]. https://arxiv.org/abs/2010.07611.

[17]	唐克, 魏飞鸣, 李东瀛, 等基于改进YOLOv8的轻量化无人机图像目标检测算法[J]. 计算机工程, 2026, 52 (3): 97- 106 Tang Ke, Wei Feiming, Li Dongying, et al Lightweight target detection algorithm for UAV images based on improved YOLOv8[J]. Computer Engineering, 2026, 52 (3): 97- 106

[18]	Tan M, Le Q V. EfficientNet: rethinking model scaling for convolutional neural networks [PP/OL]. arXiv (2019-05-28)[2025-09-15]. https://arxiv.org/abs/1905.11946.

[19]	Chen Y, Yuan X, Wang J, et al YOLO-MS: rethinking multi-scale representation learning for real-time object detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025, 47 (6): 4240- 4252 doi: 10.1109/TPAMI.2025.3538473

[1]	Heng YANG,Shaohan WANG,Qing DONG,Keyuan ZHAO,Mingliang YANG. Object pose estimation based on attention enhancement cross-modal multilayer fusion network[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1872-1880.

[2]	Jiuyuan HUO,Hao JIANG,Chen CHANG. Text detection of highway traffic sign based on improved EAST[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1881-1889.

[3]	Kaijun WU,Yunqi ZHENG,Ding WEI,Haixiang YUAN. YOLOv8s based lightweight algorithm for small object detection in aerial imagery[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1912-1923.

[4]	Hongwei LI,Jianyang DING,Zhilei CHAI. Data-efficient lightweight model for real-time semantic segmentation[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1934-1941.

[5]	Li SONG,Yuanyu WAN,Mingli SONG. Graph convolutional reinforcement learning via hybrid dilated attention mechanism[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1980-1990.

[6]	Changyuan LIU,Haijian ZHAO,Haibin WU,Jiawei LIU. Lightweight traffic police gesture recognition method for autonomous driving[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(8): 1678-1685.

[7]	Minghua ZHAO,Yuxuan LYU,Jiahao LYU,Yifei CHEN,Cheng SHI,Jing HU. Video anomaly detection based on multi-scale appearance and motion fusion[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(8): 1730-1738.

[8]	Huixia LIU,Rongjing WANG,Yuxuan GAO,Meng CAO,Suying SHENG,Youpeng MA. Adaptive neural network-based deadlock-free vehicle scheduling algorithm for unsignalized intersections[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(8): 1832-1840.

[9]	Kaiwei XU,Hafiz KHIZER BIN TALIB,Yanlong CAO,Yuanping XU,Zhijie XU,Jingchun SONG. Lightweight micro-expression recognition based on optical flow and convolutional vision Transformer[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(7): 1381-1391.

[10]	Ping WANG,Anzhi XU,Hongli ZHAO,Xiaoyuan WEI,Fulong YANG. Real-time debris flow detection method combining fluid automatic annotation and lightweight YOLOv8n[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(7): 1416-1426.

[11]	Yanchun YANG,Jialong LI. Multi-path collaboration-based and spatial-spectral prior-based hyperspectral and multispectral image fusion[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(7): 1427-1437.

[12]	Baijing WU,Guanghui YAN,Long MA,Wenxin CHENG,Yaning HUANG. Small-target water-floating garbage detection based on edge perception and cross-scale feature enhancement[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(7): 1452-1463.

[13]	Naizhou ZHANG,Yunchao ZHAO,Wei CAO,Xiaojian ZHANG. Image captioning generation based on multiple-view cross-modal feature fusion[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(6): 1205-1212.

[14]	Yunhong LI,Qiqi ZHANG,Jinni CHEN,Weichong CHEN,Xueping SU,Chengming LIANG. Text-to-image generation algorithm based on generative adversarial network and coordinate attention mechanism[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(6): 1213-1220.

[15]	Guoyan LI,Wei YU,Yupeng MEI,Minghui ZHANG,Xinqiang WANG. Building extraction from remote sensing images with global-local feature fusion[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(5): 1100-1108.

Viewed

Full text

Abstract

Cited

Shared

Discussed