Safety distance warning for forklift driving obstacles based on improved YOLOv12

doi:10.16265/j.cnki.issn1003-3033.2026.03.1262

Abstract

Abstract:

In order to solve the problems of high equipment price and large quantity demand in the existing forklift driving obstacle safety early warning distance measurement, a forklift driving obstacle safety distance early warning model based on image information was proposed. Firstly, based on deep learning technology, Squeeze-and-Excitation(SE) networks channel attention mechanism is introduced, and methods such as replacing the Intersection over Union(IoU) localization loss function with the Adaptive Threshold Focal Loss (ATFL)function are employed to improve the YOLOv12 algorithm for identifying obstacle targets in forklift travel. Secondly, on the basis of the improved YOLOv12 algorithm, the Kalman filter was introduced to improve the motion prediction model. And the distance detection method considering the camera pitch angle was used to accurately obtain the actual distance between different types of targets and the driving fork workshop. Thirdly, the kinematic process of forklift braking and forklift obstacle avoidance was analyzed, and the classification criteria of safe braking distance warning level and safety obstacle avoidance distance warning level were established, respectively. Finally, experiments were carried out to verify the feasibility of the safety warning distance of forklift driving obstacles based on image information. The results show that the real-time distance warning model can accurately identify obstacle targets in real-time and precisely determine the distance to obstacles within the permissible error range, enabling risk-level warning for obstacles during forklift operation.

Key words: YOLOv12, forklift operation, safety early warning distance, risk area, obstacle distance measurement, warning level

CLC Number:

X959其他

ZHOU Cheng, DAI Wenjie, WAN Shuhao, JU Likai. Safety distance warning for forklift driving obstacles based on improved YOLOv12[J]. China Safety Science Journal, 2026, 36(3): 89-97.

Figures/Tables 17

Fig.1

Fig.2

Table 1

Fig.3

Fig.4

Fig.5

Fig.6

Table 2

Fig.7

Fig.8

Fig.9

Fig.10

Fig.11

Fig.12

Fig.13

Fig.14

Fig.15

References 24

[1]	万舒昊. 叉车作业安全风险预警系统研究[D]. 南京: 南京理工大学, 2022.
	WAN Shuhao. Research on safety risk early warning system for forklift operations[D]. Nanjing: Nanjing University of Science and Technology, 2022.
[2]	STEPHEN S, ALIREZA B H, REZA H, et al. Analysis of forklift accident trends within Victorian industry[J]. Safety Science, 2013, 60:176-184. doi: 10.1016/j.ssci.2013.07.017
[3]	李千登, 王廷春, 崔靖文, 等. 基于深度学习算法的叉车危险操作行为检测[J]. 中国安全生产科学技术, 2020, 16(5):155-159.
	LI Qiandeng, WANG Tingchun, CUI Jingwen, et al. Detection on dangerous operation behavior of forklift based on deep learning algorithm[J]. Journal of Safety Science and Technology, 2020, 16(5):155-159.
[4]	TAN Wengchun, SIDHU M S. Review of RFID and IoT integration in supply chain management[J]. Operations Research Perspectives, 2022,9: DOI:10.1016/J.ORP.2022.100229.
[5]	HYEONGJUN K, TAEJOO K, WON J, et al. Multispectral benchmark dataset and baseline for forklift collision avoidance[J]. Sensors, 2022, 22(20):7 953-7 953. doi: 10.3390/s22207953
[6]	郭政良. 基于物联网的叉车安全监管系统设计与实现[D]. 济南: 山东大学, 2020.
	GUO Zhengliang. Design and implementation of forklift safety supervision system based on internet of things[D]. Jinan: Shandong University, 2020.
[7]	吕恩利, 阮清松, 刘妍华, 等. 基于激光扫描区域动态变化的智能叉车障碍物检测[J]. 农业工程学报, 2019, 35(3):67-74.
	LYU Enli, RUAN Qingsong, LIU Yanhua, et al. Intelligent forklift obstacle detection based on dynamic change of laser scanning area[J]. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35(3):67-74.
[8]	TAKAHASHI M, KOBAYASHI K, WATANABE K, et al. Development of prediction based emergency obstacle avoidance module by using LIDAR for mobile robot[C]. International Symposium on Soft Computing and Intelligent Systems, 2015:561-564.
[9]	DONGSHENG B, PEIKANG W. Vehicle distance detection based on monocular vision[C]. Proceedings of the 2016 IEEE International Conference on Progress in Informatics and Computing(PIC), 2016:224-228.
[10]	SONG Feng, YUBIN Qian, YAN Wang. Collision avoidance method of autonomous vehicle based on improved artificial potential field algorithm[J]. Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 2021, 235(14):3 416-3 430. doi: 10.1177/09544070211014319
[11]	ZHANG Hongge. Intelligent collision avoidance method for motion obstacles of unmanned vehicles[J]. International Journal of Vehicle Structures & Systems, 2021, 13(4):405-410.
[12]	TIAN Yunjie, YE Qixiang, DOERMANN D. YOLOv12: attention-centric real-time object detectors[J]. arXiv Preprint arXiv:2502.12524, 2025:DOI:10.48550/arXiv.1710.09412.
[13]	HU Jie, SHEN Li, SUN Gang. Squeeze-and-excitation networks[C]. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018:7 132-7 141.
[14]	LYU Wenyu, ZHAO Yian, XU Shangliang, et al. DETRs beat YOLOs on real-time object detection[J]. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Seattle, 2024: 16 965-16 974.
[15]	LIU Minggao, WANG Haifeng, DU Luyao, et al. Bearing-DETR: a lightweight deep learning model for bearing defect detection based on RT-DETR[J]. Sensors, 2024, 24(13): DOI: 10.3390/s24134262.
[16]	罗会兰, 陈鸿坤. 基于深度学习的目标检测研究综述[J]. 电子学报, 2020, 48(6):1 230-1 239.
	LUO Huilan, CHEN Hongkun. A review of research on object detection based on deep learning[J]. Acta Sinica, 2020, 48(6):1 230-1 239.
[17]	YAN Weihao. Threshold-adaptive unsupervised focal loss for domain adaptation in semantic segmentation[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(1): 752-763. doi: 10.1109/TITS.2022.3210759
[18]	ZEHUAN Jing, PENG Li, BIN Wu, et al. An adaptive focal loss function based on transfer learning for few-shot radar signal intra-pulse modulation classification[J]. Remote Sensing, 2022, 14(8):1 950-1 950. doi: 10.3390/rs14081950
[19]	EVERINGHAM M, GOOL L V, WILLIAMS C K I, et al. The pascal visual object classes (VOC) challenge[J]. International Journal of Computer Vision, 2010, 88(2):303-338. doi: 10.1007/s11263-009-0275-4
[20]	ZHANG Hongyi, CISSE M, DAUPHIN Y N, et al. MixUp: beyond empirical risk minimization[J]. arXiv preprint, 2018:DOI:10.48550/arXiv.1710.09412.
[21]	YUN Sangdoo, HAN Dongyoon, OH S J, et al. CutMix: regularization strategy to train strong classifiers with localizable features[C]. IEEE/CVF International Conference on Computer Vision (ICCV). Proceedings, 2019:6 022-6 031.
[22]	苑振磊. 基于深度学习的视频中运动目标测距技术研究[D]. 深圳: 深圳大学, 2023.
	YUAN Zhenlei. Research on moving target ranging technology in video based on deep learning[D]. Shenzhen: Shenzhen University, 2023.
[23]	ALEX B, GE Zongyuan, LIONEL O, et al. Simple online and realtime tracking[C]. 2016 IEEE International Conference on Image Processing (ICIP), 2016:3 464-3 468.
[24]	NICOLAI W, ALEX B, DIETRICH P. Simple online and realtime tracking with a deep association metric[C]. 2017 IEEE International Conference on Image Processing (ICIP), 2017:3 645-3 649.

模型	mAP/%	FPS/(帧·s^-1)
YOLOv12	94.9	227.01
SSD	85.2	130.2
Faster-RCNN	93.6	110.4
YOLOv12-CRA	96.1	197.07

实际距离	检测距离
2	2.0
5	5.1
10	9.8
15	14.5