基于空间域和频率域方法的烟雾检测

doi:10.3969/j.issn.1000-5641.2023.05.013

摘要/Abstract

摘要：

变电站等工业场景中, 基于监控视频的视觉烟雾检测已被作为一种新的环境辅控方式, 用于辅助或代替基于烟雾传感器的烟雾检测. 但是, 工业应用中要求视觉烟雾检测算法在保证误检率低的基础上, 要尽可能降低漏检率. 针对该问题, 基于空间域和频率域方法, 提出了一种新的烟雾检测算法, 分别在空间域和频率域进行烟雾检测: 在空间域上, 在提取烟雾运动特性的基础上, 设计了提取烟雾“蒙版特性”的方法, 以保证较低的漏检率; 在频率域上, 分别结合滤波模块和神经网络模块, 以进一步降低误检率. 最后通过融合后处理策略, 得到最终检测结果, 从而平衡漏检率和误检率. 在测试数据集上, 所提烟雾检测算法的误检率达到了0.053, 漏检率达到了0.113, 实现了误检率和漏检率的良好平衡. 所提检测方法适用于变电站等实际工业场景的烟雾检测.

关键词: 视觉烟雾检测, 空间域和频率域, 神经网络, 三维傅里叶变换

Abstract:

In industrial scenarios such as substations, video-based visual smoke detection has been adopted as a new environmental monitoring method to assist or replace smoke sensors. However, in industrial applications, visual smoke detection algorithms are required to maintain a low false detection rate while minimizing the missed detection rate. To address this, this study proposes a smoke detection algorithm based on spatial and frequency domain methods, which perform smoke detection in both domains. In the spatial domain, in addition to extracting smoke motion characteristics, this study designed a method for extracting smoke mask characteristics, which effectively ensures a low missed detection rate. In the frequency domain, this study combined filtering and neural network modules to further reduce the false detection rate. Finally, a fusion domain post-processing strategy was designed to obtain the final detection results. In experiments conducted on a test dataset, the smoke detection algorithm achieved a false detection rate of 0.053 and missed detection rate of 0.113, demonstrating a good balance between false alarms and missed detections, which is suitable for smoke detection in substation industrial scenes.

Key words: visual smoke detection, spatial and frequency domains, neural network, three-dimensional Fourier transform

中图分类号:

TP391

盛连军, 汤致轩, 茅晓亮, 白帆, 黄定江. 基于空间域和频率域方法的烟雾检测[J]. 华东师范大学学报（自然科学版）, 2023, 2023(5): 147-163.

Lianjun SHENG, Zhixuan TANG, Xiaoliang MAO, Fan BAI, Dingjiang HUANG. Smoke detection based on spatial and frequency domain methods[J]. Journal of East China Normal University(Natural Science), 2023, 2023(5): 147-163.

图/表 15

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

表1

表2

表3

图12

参考文献 27

1	李孟超, 王允平, 李献伟, 等.. 智能变电站及技术特点分析. 电力系统保护与控制, 2010, 38 (18): 59- 62,79.
2	唐慧明, 张健.. 变电站远程图像监控系统设计. 电力系统自动化, 2001, (7): 56- 58.
3	宋伟, 王玉平, 张文杰, 等.. 基于颜色与运动特征的变电站实时烟雾检测. 传感器与微系统, 2015, 34 (12): 127- 130. doi: 10.13873/J.1000-9787(2015)12-0127-04
4	TÖREYIN B U, DEDEOĞLU Y, CETIN A E. Wavelet based real-time smoke detection in video [C]// 2005 13th European Signal Processing Conference. IEEE, 2005. https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7077943.
5	YU C Y, ZHANG Y M, FANG J, et al. Texture analysis of smoke for real-time fire detection [C]// 2009 2nd International Workshop on Computer Science and Engineering. IEEE, 2009: 511-515. DOI: 10.1109/WCSE.2009.864
6	ALAMGIR N, NGUYEN K, CHANDRAN V, et al.. Combining multi-channel color space with local binary co-occurrence feature descriptors for accurate smoke detection from surveillance videos. Fire Safety Journal, 2018, 102, 1- 10.
7	FRIZZI S, KAABI R, BOUCHOUICHA M, et al. Convolutional neural network for video fire and smoke detection [C]// IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society. IEEE, 2016: 877-882.
8	赵丽宏, 付雪, 肖博宇, 等.. 基于卷积神经网络的森林火灾烟雾探测算法研究. 中国新技术新产品, 2019, (12): 23- 24.
9	GAGLIARDI A, DE GIOIA F, SAPONARA S.. A real-time video smoke detection algorithm based on Kalman filter and CNN. Journal of Real-Time Image Processing, 2021, 18, 2085- 2095.
10	VALIKHUJAEV Y, ABDUSALOMOV A, CHO Y I.. Automatic fire and smoke detection method for surveillance systems based on dilated CNNs. Atmosphere, 2020, 11 (11): 1241. doi: 10.3390/atmos11111241
11	WU Y L, CHEN M H, WO Y, et al.. Video smoke detection base on dense optical flow and convolutional neural network. Multimedia Tools and Applications, 2021, 80, 35887- 35901.
12	MARDANI K, VRETOS N, DARAS P.. Transformer-based fire detection in videos. Sensors, 2023, 23 (6): 3035. doi: 10.3390/s23063035
13	ALMEIDA J S, HUANG C, NOGUEIRA F G, et al.. EdgeFireSmoke: A novel lightweight cnn model for real-time video fire–smoke detection. IEEE Transactions on Industrial Informatics, 2022, 18 (11): 7889- 7898.
14	CHEN X, XUE Y P, ZHU Y L, et al.. A novel smoke detection algorithm based on improved mixed Gaussian and YOLOv5 for textile workshop environments. IET Image Processing, 2023, 17, 1991- 2004. doi: 10.1049/ipr2.12719
15	DE VENÂNCIO P V A B, CAMPOS R J, REZENDE T M, et al.. A hybrid method for fire detection based on spatial and temporal patterns. Neural Computing and Applications, 2023, 35 (13): 9349- 9361.
16	WU Z S, XUE R, LI H.. Real-time video fire detection via modified YOLOv5 network model. Fire Technology, 2022, 58, 2377- 2403.
17	YUAN F N, SHI Y, ZHANG L, et al.. A cross-scale mixed attention network for smoke segmentation. Digital Signal Processing, 2023, 134, 103924.
18	AHN Y, CHOI H, KIM B S.. Development of early fire detection model for buildings using computer vision-based CCTV. Journal of Building Engineering, 2023, 65, 105647.
19	HUO Y N, ZHANG Q X, JIA Y, et al.. A deep separable convolutional neural network for multiscale image-based smoke detection. Fire Technology, 2022, 58, 1445- 1468.
20	PICCARDI M. Background subtraction techniques: A review [C]// 2004 IEEE International Conference on Systems, Man and Cybernetics (IEEE Cat. No. 04CH37583). IEEE, 2004: 3099-3104.
21	FARNEBÄCK G. Two-frame motion estimation based on polynomial expansion [C]// Image Analysis, SCIA 2003, Lecture Notes in Computer Science, vol 2749. Berlin: Springer, 2003: 363-370.
22	SINGLA N.. Motion detection based on frame difference method. International Journal of Information & Computation Technology, 2014, 4 (15): 1559- 1565.
23	COOLEY JW, LEWIS PA, WELCH PD.. The fast Fourier transform and its applications. IEEE Transactions on Education, 1969, 12 (1): 27- 34.
24	AHMED N, NATARAJAN T, RAO K R.. Discrete cosine transform. IEEE Transactions on Computers, 1974, C-23 (1): 90- 93.
25	ZHANG D S. Fundamentals of Image Data Mining: Analysis, Features, Classification and Retrieval [M]. 2nd ed. Cham : Springer , 2019: 45-54.
26	JI S W, XU W, YANG M, et al.. 3D convolutional neural networks for human action recognition. IEEE Transactions on Tattern Analysis and Machine Intelligence, 2013, 35 (1): 221- 231.
27	YUAN F N. Video smoke detection [EB/OL]. (2011-08-28)[2023-04-03]. http://staff.ustc.edu.cn/ ~ yfn/vsd.html.

算法	误检率	漏检率
基线算法	0.776	0.013
基线算法 + 蒙版特性	0.671	0.025
基线算法 + 蒙版特性 + 频域滤波	0.066	0.088
基线算法 + 蒙版特性 + 频域滤波 + 三维神经网络	0.053	0.113

算法	误检率	漏检率
基于小波变换的烟雾检测算法^[4]	0.115	0.184
基于纹理分析的烟雾检测算法^[5]	0.092	0.106
基于改进YOLOv5的烟雾检测算法^[16]	0.079	0.173
本文算法	0.053	0.113

算法	单张图片平均推理时间/ms
算法	在GPU上	在CPU上	在NPU上
基于小波变换的烟雾检测算法^[4]	42	50	–
基于纹理分析的烟雾检测算法^[5]	37	42	–
基于改进YOLOv5的烟雾检测算法^[16]	55	170	68
本文算法	48	161	65

[1]	黄彩蝶, 王昕萍, 陈良育, 刘勇. 基于堆叠门控循环单元残差网络的知识追踪模型研究[J]. 华东师范大学学报（自然科学版）, 2022, 2022(6): 68-78.
[2]	吴豪杰, 王妍洁, 蔡文炳, 王飞, 刘洋, 蒲鹏, 林绍辉. 基于隐层相关联算子的知识蒸馏方法[J]. 华东师范大学学报（自然科学版）, 2022, 2022(5): 115-125.
[3]	李继洲, 林欣. 基于递归结构的神经网络架构搜索算法[J]. 华东师范大学学报（自然科学版）, 2022, 2022(4): 31-42.
[4]	赵晓臻, 童卫青, 刘咏梅. 一种面向服装样板图分类的图卷积神经网络[J]. 华东师范大学学报（自然科学版）, 2022, 2022(4): 56-66.
[5]	王诺, 李丽颖, 钱栋炜, 魏同权. 基于网络压缩与切割的深度模型边云协同加速机制研究[J]. 华东师范大学学报（自然科学版）, 2021, 2021(6): 112-123.
[6]	刘波, 白晓东, 张更新, 沈俊, 谢继东, 赵来定, 洪涛. 深度学习在认知无线电中的应用研究综述[J]. 华东师范大学学报（自然科学版）, 2021, 2021(1): 36-52.
[7]	赖叶静, 郝珊锋, 黄定江. 深度神经网络模型压缩方法与进展[J]. 华东师范大学学报（自然科学版）, 2020, 2020(5): 68-82.
[8]	周晓旭, 刘迎风, 付英男, 朱仁煜, 高明. 网络顶点表示学习方法[J]. 华东师范大学学报（自然科学版）, 2020, 2020(5): 83-94.
[9]	刘锦, 赵维锐. 具有时变时滞的中立型神经网络的稳定性分析[J]. 华东师范大学学报（自然科学版）, 2020, 2020(4): 35-44.
[10]	贺小娟, 郭新顺. 基于特征优化的广告点击率预测模型研究[J]. 华东师范大学学报（自然科学版）, 2020, 2020(4): 147-155.
[11]	邵明锐, 马登豪, 陈跃国, 覃雄派, 杜小勇. 基于社区问答数据迁移学习的FAQ问答模型研究[J]. 华东师范大学学报(自然科学版), 2019, 2019(5): 74-84.
[12]	赵波, 田秀霞, 李灿. 基于自适应神经网络的电网稳定性预测[J]. 华东师范大学学报(自然科学版), 2019, 2019(5): 133-142.
[13]	袁培森, 张勇, 李美玲, 顾兴健. 基于深度哈希学习的商标图像检索研究[J]. 华东师范大学学报(自然科学版), 2018, 2018(5): 172-182.
[14]	白玫, 吴辉. 利用SOM神经网络研究长江口邻近海域海表温度特征[J]. 华东师范大学学报(自然科学版), 2018, 2018(4): 184-194.
[15]	郁可人, 傅云斌, 董启文. 基于神经网络语言模型的分布式词向量研究进展[J]. 华东师范大学学报(自然科学版), 2017, 2017(5): 52-65,79.