Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)  2020, Vol. 54 Issue (12): 2430-2436    DOI: 10.3785/j.issn.1008-973X.2020.12.018
计算机与控制工程     
采用Anchor-Free网络结构的实时火灾检测算法
晋耀(),张为*()
天津大学 微电子学院，天津 300072
Real-time fire detection algorithm with Anchor-Free network architecture
Yao JIN(),Wei ZHANG*()
School of Microelectronics, Tianjin University, Tianjin 300072, China
 全文: PDF(1682 KB)   HTML
摘要：

为了解决现有的火灾检测算法中模型复杂，实时性差，检测精度较低的问题，提出快速高效的火灾检测算法. 该算法采用Anchor-Free网络结构，克服了Anchor方法中超参数过多、网络结构复杂的缺点；选用MobileNetV2作为基础特征提取网络，满足了检测的高实时性需求；针对火焰目标的形状、尺度多变的特点，设计适合于火焰检测的分类与边框预测子网络；引入特征选择模块，在特征金字塔网络中自动选择更合适的金字塔特征层. 算法在自建数据集上的检测精度达到90.1%；在公开数据集上取得了较好的检测结果，其检测速度可达24.6 帧/s. 实验结果表明，算法的网络模型简单，检测精度较高，检测速度较快；综合性能优于现有的其他火灾检测算法.
关键词： 计算机视觉;  目标检测;  火灾检测;  Anchor-Free;  可形变卷积;  特征选择模块    
Abstract:

The existing fire detection algorithms had problems such as too complicated model structure, poor real-time ability, and low detection rate. A fast and efficient fire detection algorithm based on convolutional neural network was proposed, in order to solve those problems. The Anchor-Free network structure was adoped as the whole architecture, which avoided the shortcomings of too many hyper-parameters and complex network structure in the Anchor-based detector. MobileNetV2 was selected as the feature extraction backbone network in order to meet the requirements of high real-time ability. A classification and a box prediction task-specific subnet suitable for fire detection were designed considering that the size and shape of the flame were different from other objects. A feature selective module was introduced to help the feature pyramid network automatically selected a more suitable pyramid feature layer. The algorithm achieves 90.1% accuracy on the self-built dataset, and has good performance on the public fire dataset. The speed of the network achieves 24.6 frames per second. The experimental results show that the detection precision of the proposed algorithm is high and the speed is fast while being simpler. The comprehensive performance of the algorithm is better than other existing fire detection algorithms.
Key words: computer vision    object detection    fire detection    Anchor-Free    deformable convolution    feature selective module
收稿日期: 2019-12-18 出版日期: 2020-12-31
CLC:  TP 391.4  
基金资助: 公安部技术研究计划资助项目（2017JSYJC35）
通讯作者: 张为     E-mail: jin_yao@tju.edu.cn;tjuzhangwei@tju.edu.cn
作者简介: 晋耀（1996—），男，硕士生，从事数字图像处理、模式识别研究. orcid.org/0000-0002-2586-2921. E-mail： jin_yao@tju.edu.cn
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　　
晋耀
张为

引用本文:

晋耀,张为. 采用Anchor-Free网络结构的实时火灾检测算法[J]. 浙江大学学报(工学版), 2020, 54(12): 2430-2436.

Yao JIN,Wei ZHANG. Real-time fire detection algorithm with Anchor-Free network architecture. Journal of ZheJiang University (Engineering Science), 2020, 54(12): 2430-2436.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2020.12.018    或    http://www.zjujournals.com/eng/CN/Y2020/V54/I12/2430

图 1  采用Anchor-Free网络结构的火灾检测算法网络模型
图 2  特征金字塔层中的有效区域
图 3  FPN结构中特征选择过程
图 4  特征选择模块
算法 可形变卷积
模块 		特征选择
模块 		AP50 /% 单张图片检测
时间/ms
RetinaNet-
MobileNet 		? ? 87.3 41.5
本文算法 ? √ 86.3 35.7
√ ? 87.9 37.1
√ √ 90.1 40.7
表 1  消融实验结果
图 5  RetinaNet-MobileNetV2与Anchor-Free网络结构检测结果对比
算法 主干网络 AP50 /% 单张图片检测时间/ms
YoloV3 DarkNet-53 89.2 38.4
SSD300 Vgg[21] 85.0 27.6
ReinaNet-MobileNet MobileNetV2 87.3 41.5
本文 MobileNetV2 90.1 40.7
表 2  经典目标检测算法与本文算法的检测结果对比
图 6  火灾检测公共数据集
算法 TPR/% TNR/%
图6（a） 图6（b） 图6（c） 图6（d） 图6（e） 图6（f） 图6（g） 图6（h）
文献[22] 94.59 ? 92.25 ? 95.45 ? ? ?
文献[23] 94.98 ? 93.00 ? 96.50 ? ? ?
文献[24] 91.20 ? ? 88.60 95.80 ? ? ?
文献[25] 95.70 ? 93.10 ? 100 ? ? ?
文献[26] 93.00 ? 91.00 ? 100 ? ? 100
文献[2] 100 94.29 96.15 ? 100 95.20 ? ?
本文 100 96.15 94.75 97.95 99.49 97.41 100 100
表 3  现有火灾检测算法与本文算法的检测结果对比
1 SEEBAMRUNGSAT J, PRAISING S, RIYAMONGKOL P. Fire detection in the buildings using image processing [C]// 2014 3rd ICT International Student Project Conference. Thailand: IEEE, 2014: 95-98.
2 HASHEMZADEH M, ZADEMEHDI A Fire detection for video surveillance applications using ICA K-medoids-based color model and efficient spatio-temporal visual features[J]. Expert Systems with Applications, 2019, 130: 60- 78
doi: 10.1016/j.eswa.2019.04.019
3 KIM B, LEE J A video-based fire detection using deep learning models[J]. Applied Sciences, 2019, 9 (14): 2862
doi: 10.3390/app9142862
4 WU S, ZHANG L. Using popular object detection methods for real time forest fire detection [C]// 2018 11th International Symposium on Computational Intelligence and Design. Hangzhou: IEEE, 2018: 280-284.
5 LAW H, DENG J. CornerNet: detecting objects as paired keypoints [C]// Proceedings of the European Conference on Computer Vision. Munich: Springer, 2018: 734-750.
6 ZHU C, HE Y, SAVVIDES M. Feature selective anchor-free module for single-shot object detection [C]// Proceedings of the IEEE International Conference on Computer Vision. Long Beach: IEEE, 2019: 840-849.
7 TIAN Z, SHEN C, CHEN H, et al. FCOS: fully convolutional one-stage object detection [C]// Proceedings of the IEEE International Conference on Computer Vision. Seoul: IEEE, 2019: 9627-9636.
8 KONG T, SUN F, LIU H, et al. FoveaBox: beyond anchor-based object detector [EB/OL]. [2019-04-08]. https://arxiv.org/ftp/arxiv/papers/1904/1904.03797.pdf.
9 SANDLER M, HOWARD A, ZHU M, et al. MobileNetV2: inverted residuals and linear bottlenecks [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 4510-4520.
10 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: 2980-2988.
11 LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 2117-2125.
12 DAI J, QI H, XIONG Y, et al. Deformable convolutional networks [C]// Proceedings of the IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 764-773.
13 SUN F, KONG T, HUANG W, et al Feature pyramid reconfiguration with consistent loss for object detection[J]. IEEE Transactions on Image Processing, 2019, 28 (10): 5041- 5051
doi: 10.1109/TIP.2019.2917781
14 NAIR V, HINTON G E. Rectified linear units improve restricted boltzmann machines [C]// Proceedings of the 27th International Conference on Machine Learning. Haifa: ACM, 2010: 807-814.
15 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
16 LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft coco: common objects in context [C]// European Conference on Computer Vision. Swizterland: Springer, 2014: 740-755.
17 REDMON J, FARHADI A. Yolov3: an incremental improvement [EB/OL]. [2018-04-08]. https://arxiv.org/pdf/1804.02767.pdf.
18 LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector [C]// European Conference on Computer Vision. Amsterdam: Springer, 2016: 21-37.
19 HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
20 NEWELL A, YANG K, DENG J. Stacked hourglass networks for human pose estimation [C]// European Conference on Computer Vision. Amsterdam: Springer, 2016: 483-499.
21 SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition [C]// International Conference on Learning Representations. San Diego: ICLR, 2015: 1–14.
22 BORGES P V K, IZQUIERDO E A probabilistic approach for vision-based fire detection in videos[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2010, 20 (5): 721- 731
doi: 10.1109/TCSVT.2010.2045813
23 TRUONG T X, KIM J M Fire flame detection in video sequences using multi-stage pattern recognition techniques[J]. Engineering Applications of Artificial Intelligence, 2012, 25 (7): 1365- 1372
doi: 10.1016/j.engappai.2012.05.007
24 吴茜茵, 严云洋, 杜静, 等 多特征融合的火焰检测算法[J]. 智能系统学报, 2015, 10 (2): 240- 247
WU Xi-yin, YAN Yun-yang, DU Jing, et al Fire detection based on fusion of multiple features[J]. CAAI Transactions on Intelligent Systems, 2015, 10 (2): 240- 247
25 KONG S G, JIN D, LI S, et al Fast fire flame detection in surveillance video using logistic regression and temporal smoothing[J]. Fire Safety Journal, 2016, 79: 37- 43
doi: 10.1016/j.firesaf.2015.11.015
26 梅建军, 张为 基于ViBe与机器学习的早期火灾检测算法[J]. 光学学报, 2018, 38 (7): 60- 67
MEI Jian-jun, ZHANG Wei Early fire detection algorithm based on ViBe and machine learning[J]. Acta Optica Sinica, 2018, 38 (7): 60- 67
[1] 徐利锋,黄海帆,丁维龙,范玉雷. 基于改进DenseNet的水果小目标检测[J]. 浙江大学学报(工学版), 2021, 55(2): 377-385.
[2] 郑浦,白宏阳,李伟,郭宏伟. 复杂背景下的小目标检测算法[J]. 浙江大学学报(工学版), 2020, 54(9): 1777-1784.
[3] 张峻宁,苏群星,刘鹏远,王正军,谷宏强. 基于空间约束的自适应单目3D物体检测算法[J]. 浙江大学学报(工学版), 2020, 54(6): 1138-1146.
[4] 郑晨斌,张勇,胡杭,吴颖睿,黄广靖. 目标检测强化上下文模型[J]. 浙江大学学报(工学版), 2020, 54(3): 529-539.
[5] 林志洁,罗壮,赵磊,鲁东明. 特征金字塔多尺度全卷积目标检测算法[J]. 浙江大学学报(工学版), 2019, 53(3): 533-540.
[6] 叶芳芳,许力. 实时的静止目标与鬼影检测及判别方法[J]. 浙江大学学报(工学版), 2015, 49(1): 181-185.
[7] 刘辉涛,汪李明,李建龙. 声纳强脉冲干扰的自适应抵消方法[J]. J4, 2011, 45(3): 515-519.
[8] 刘士荣, 王凯, 邱雪娜. 基于自适应混合高斯模型全方位视觉目标检测[J]. J4, 2010, 44(7): 1387-1393.
[9] 潘翔 马德强 吴贻军 张光富 姜哲圣. 基于视觉着陆的无人机俯仰角与高度估计[J]. , 2009, 43(4): 692-696.
[10] 刘佳 于慧敏. 基于水平集的运动目标检测和速度估算[J]. J4, 2009, 43(2): 244-249.
[11] 赵云峰 陈隆道 孟庆利. 提高红外经纬仪跟踪弱小目标精度的新算法[J]. J4, 2008, 42(7): 1169-1173.
[12] 楼斌 沈海斌 严晓浪. 基于运动目标检测的自嵌入视频水印[J]. J4, 2008, 42(3): 382-386.
[13] 漆随平 张宏建 骆志坚 周洪亮. 虚拟多传感器信息融合的在线钢材视觉检测[J]. J4, 2005, 39(9): 1363-1367.