复杂环境下小尺度烟火目标检测研究
作者:
作者单位:

1.南京工程学院 自动化学院;2.江苏省智能感知技术与装备工程研究中心

中图分类号:

TP391

基金项目:

国家自然科学基金面上项目(51675259);江苏省智能感知技术与装备工程研究中心开放基金项目(ITS202103)


Research on small scale fireworks target detection in complex environment
Author:
Affiliation:

1.School of Automation,Nanjing Institute of technology,Nanjing,Jiangsu;2.Province Research Center of Intellisense Technology and System,Nanjing,Jiangsu

  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献
  • | |
  • 引证文献
  • | |
  • 文章评论
    摘要:

    针对复杂环境下起火点目标尺寸较小、起火点特征易与实际场景混淆导致烟火检测效率及准确率低等问题,提出了一种基于改进YOLOv5的小尺度烟火目标检测方法。首先,在原始YOLOv5模型输出的第三个检测层上增加第四个检测层,以此获取更大的特征图对小目标进行检测,加强网络模型的特征提取能力。其次,为解决目标在被遮挡的场景中容易出现漏检的问题,将原网络中用于计算目标框损失函数的 GIOU_Loss 替换成 DIOU_Loss。最后,利用TensorRT对模型进行压缩和加速优化,并将其部署到Jetson TX2开发板上进行加速推理实验,通过复制增强方法扩充实际烟火场景数据,大量实验结果表明,提出方法用于复杂环境下的小尺度烟火目标检测不仅检测速度快而且精度高,适于推广应用。

    Abstract:

    Aiming at the problem of low efficiency and accuracy of fireworks detection due to the small size of fire target and the confusion of fire feature with actual scene in complex environment, a small scale fireworks target detection method based on improved YOLOv5 is proposed. Firstly, a fourth detection layer is added to the third detection layer output in the original YOLOv5 model, so as to obtain a larger feature map for small target detection and strengthen the feature extraction capability of the network model. Secondly, in order to solve the problem that the target is prone to miss detection in the shielded scene, GIOU_Loss used to calculate the regression loss function of the target frame in the original network is replaced by DIOU_Loss. Finally, TensorRT is used to compress and accelerate the optimization of the model, and it is deployed to the Jetson TX2 development board for accelerated reasoning experiments. More fireworks scene data are constructed by replication enhancement method and a large number of experimental results show that the proposed method not only has a fast convergence speed, but also has a higher accuracy for small scale fire spot detection. It is suitable for popularization and application.

    参考文献
    [1] 李继超, 郭聖煜, 孔刘林, 等.施工现场火焰检测和预警机器人设计及应用[J]. 中国安全科学学报, 2021, 31(04): 141-146.I Jichao, GUO Shengyu, KONG Liulin, et al. Design and application of flame detection and early warning robot in construction site[J]. Chinese Journal of Safety Science, 2021, 31(04): 141-146.
    [2] 胡凯,郑翡,卢飞宇,等.基于深度学习的行为识别算法综述[J].南京信息工程大学学报(自然科学版),2021,13(06):730-743.U Kai, ZHENG Fei, LU Feiyu, et al. A Review of Behavior Recognition Algorithms based on deep Learning [J]. Journal of Nanjing University of Information Science Technology (Natural Science Edition),2021,12(06): 730-743.
    [3] 张学颖, 杨遂军, 傅琳, 等. 基于组合传感器阵列的广谱火焰高度检测[J]. 仪表技术与传感器, 2013(06): 93-95.HANG Xueying, YANG Suijun, FU Lin, et al. Broad-spectrum flame height detection based on combined sensor array[J]. Instrumentation Technology and Sensors, 2013(06): 93-95.
    [4] 张驰, 孟庆浩, 井涛. 基于改进GMM和多特征融合的视频火焰检测算法[J]. 激光与光电子学进展, 2021, 58(04): 136-145.HANG Chi, MENG Qinghao, JIN Tao. Video flame detection algorithm based on improved GMM and multi-feature fusion[J]. Advances in Lasers and Optoelectronics, 2021, 58(04): 136-145.
    [5] 陈西江,安庆,班亚.优化EfficientDet深度学习的车辆检测[J].南京信息工程大学学报(自然科学版),2021,13(06):653-660.HEN Xijiang, AN Qing, BAN Ya. Optimization of EfficientDet Deep Learning for Vehicle Detection [J]. Journal of Nanjing University of Information Science Technology (Natural Science Edition), 2021, 13(06):653-660.
    [6] 陈浩霖,高尚兵,相林,等.FIRE-DET: 一种高效的火焰检测模型[J/OL].南京信息工程大学学报(自然科学版):1-11[2022-07-09].HEN Haolin, GAO Shangbing,XIANG Lin, at al. FIRE DET: An efficient flame detection model [J]. Journal of Nanjing University of Information Science Technology (Natural Science Edition):1-11[2022-07-09].
    [7] 石磊, 张海刚, 杨金锋. 基于改进型SSD的视频烟火检测算法[J]. 计算机应用与软件, 2021, 38(12): 161-167+173.HI Lei, ZHANG Haigang, YANG Jinfeng. Video fireworks detection algorithm based on improved SSD[J]. Computer Applications and Software, 2021, 38(12): 161-167+173.
    [8] 李欣健, 张大胜, 孙利雷, 等. 复杂场景下基于CNN的轻量火焰检测方法[J]. 模式识别与人工智能, 2021, 34(05): 415-422.I Xinjian, ZHANG Dasheng, SUN Lilei, et al. CNN-based lightweight flame detection method in complex scenes[J]. Pattern Recognition and Artificial Intelligence, 2021, 34(05): 415-422.
    [9] 赵媛媛, 朱军, 谢亚坤, 等.改进Yolo-v3的视频图像火焰实时检测算法[J]. 武汉大学学报(信息科学版), 2021, 46(03): 326-334.HAO Yuanyuan, ZHU Jun, XIE Yakun, et al. Improved Yolo-v3 video image flame real-time detection algorithm[J]. Journal of Wuhan University (Information Science Edition), 2021, 46(03): 326-334.
    [10] Shorten C, Khoshgoftaar T M. A survey on Image Data Augmentation for Deep Learning[J]. Journal of Big Data, 2019, 6(1): 1-48.
    [11] Bai Y, Zhang Y, Ding M, et al. SOD-MTGAN: Small object detection via multi-task generative adversarial network[C]// Proceedings of the European Conference on Computer Vision (ECCV). Cham: Springer, 2018: 206-221.
    [12] Hu C, Tang P, Jin WD, et al. Real-time fire detection based on deep convolutional long-recurrent networks and optical flow method[C]. In: Wang D, eds. IEEE 37th Chinese Control Conference. Wuhan: CCC, 2018: 9061-9066.
    [13] Leng J, Ren Y, Jiang W, et al. Realize your surroundings: Exploiting context information for small object detection[J]. Neurocomputing, 2021, 433: 287-299.
    [14] Lin T Y, Dollar P, Girshick R, et al. Feature pyramid networks for object detection[C]. In: Mortensen E, eds. IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: CVPR, 2017: 2117-2125.
    [15] Du S, Zhang P, Zhang B, et al. Weak and Occluded Vehicle Detection in Complex Infrared Environment Based on Improved YOLOv4[J]. IEEE Access, 2021, PP(99): 1-1.
    [16] Zhao Z Q, Zheng P, Xu S, et al. Object detection with deep learning: A review[J]. IEEE Transactions on Neural Networks and Learning Systems, 2019, 30(11): 3212-3232.
    [17] Pan S J, Qiang Y. A Survey on Transfer Learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(10): 1345-1359.
    [18] Tao L, Hong T, Guo Y C, et al. Drone identification based on CenterNet-TensorRT[C]. 2020 IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB), 2020: 1-5.
    [19] Min J, Lee Y. An Experimental View on Fairness between HTTP/1. and HTTP/2[C]. 2019 International Conference on Information Networking (ICOIN), 2019: 399-401.
    [20] Wang X H, Yue X B, Li H Y, et al. A high-efficiency dirty-egg detection system based on YOLOv4 and TensorRT[C]. 2021 International Conference on Advanced Mechatronic Systems (ICAMechS), 2021: 75-80.
    [21] Qiang Z, Yuan Y W, Liang Z, et al. Research on Real-Time Reasoning Based on JetSon TX2 Heterogeneous Acceleration YOLOv4[C]// 2021 IEEE 6th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA), 2021: 455-459.
    相似文献
    引证文献
    引证文献 [0]
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

温秀兰,焦良葆,李子康,姚波,唐国寅.复杂环境下小尺度烟火目标检测研究[J].南京信息工程大学学报,,():

复制
分享
文章指标
  • 点击次数:421
  • 下载次数: 0
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 收稿日期:2022-07-10
  • 最后修改日期:2022-07-25
  • 录用日期:2022-08-16

地址:江苏省南京市宁六路219号    邮编:210044

联系电话:025-58731025    E-mail:nxdxb@nuist.edu.cn

南京信息工程大学学报 ® 2025 版权所有  技术支持:北京勤云科技发展有限公司