PU Jianfei,WEI Wei,WU Diyong,et al.Video Smoke Detection based on Smoke Area and Lightweight Model[J].Journal of Chengdu University of Information Technology,2023,38(03):281-290.[doi:10.16836/j.cnki.jcuit.2023.03.006]
基于烟雾区域和轻量化模型的视频烟雾检测
- Title:
- Video Smoke Detection based on Smoke Area and Lightweight Model
- 文章编号:
- 2096-1618(2023)03-0281-10
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 烟雾是早期火灾发生的典型特征,针对烟雾的智能检测能有效降低森林火灾造成的破坏。为了对监控视频中烟雾的及早检测,提出一种基于烟雾区域及轻量化模型的烟雾检测算法。首先通过残差帧堆叠获取视频中的运动区域,然后再利用自适应暗通道掩码对运动区域进一步筛选获得疑似烟雾块。在此基础上,又设计了一个轻量化的神经网络模型用于烟雾识别,模型利用卷积局部感知的特性提取烟雾的浅层特征,而在网络深层则将卷积和self-attention相结合,通过比较全局相似度,在浅层特征图的基础上获取烟雾的全局信息。实验结果表明,算法具有较强的鲁棒性,无论是远距离烟雾还是近距离烟雾均有良好的检测效果。
- Abstract:
- Smoke is a typical feature of early fires, and intelligent detection of smoke can effectively reduce the damage caused by forest fires. In order to achieve early detection of smoke in surveillance video, this paper proposes a smoke detection algorithm based on the smoke area and lightweight model. The algorithm first obtains the motion regions in the video by stacking the residual frames,and then uses the adaptive dark channel mask to further screen the motion regions to obtain suspected smoke blocks. On this basis, this paper proposes a lightweight neural network model for smoke recognition. This model uses the characteristics of convolutional local perception to extract the shallow features of smoke, and in the deep layer of the network, convolution and self-attention are related. The global information of smoke is obtained on the basis of shallow feature maps by comparing the global similarity. The experimental results show that the algorithm in this paper has strong robustness, and it has a good detection effect for long-distance smoke and short-distance smoke.
参考文献/References:
[1] 严云洋,高尚兵,郭志波,等.基于视频图像的火灾自动检测[J].计算机应用研究,2008(4):1075-1078.
[2] RUSSO A U,DEB K,TISTA S C,et al.Smoke Detection Method Based on LBP and SVM from Surveillance Camera[C].proceedings of the 2018 International Conference on Computer,Communication,Chemical,Material and Electronic Engineering(IC4ME2),2018.
[3] TIANTIAN T,LINHAN D,ZHIJIAN Y.Smoke Image Recognition Based on Local Binary pattern[C].proceedings of the Proceedings of the 2017 5th International Conference on Mechatronics,Materials,Chemistry and Computer Engineering(ICMMCCE 2017),2017.
[4] LIU Z,YANG X,LIU Y,et al.Smoke-Detection Framework for High-Definition Video Using Fused Spatial- and Frequency-Domain Features[J].IEEE Access,2019,7:89687-89701.
[5] GAO Y,CHENG P.Forest Fire Smoke Detection Based on Visual Smoke Root and Diffusion Model[J].Fire Technology,2019,55(5):1801-1826.
[6] WANG S,HE Y,YANG H,et al.Video smoke detection using shape,color and dynamic features[J].Journal of Intelligent & Fuzzy Systems,2017,33:305-313.
[7] ZHOU Z,SHI Y,GAO Z,et al.Wildfire smoke detection based on local extremal region segmentation and surveillance[J].Fire Safety Journal,2016,85:50-58.
[8] WANG L,LI A.Early fire recognition based on multi-feature fusion of video smoke[C].proceedings of the 2017 36th Chinese Control Conference(CCC),2017.
[9] ZHANG F,QIN W,LIU Y,et al.A Dual-Channel convolution neural network for image smoke detection[J].Multimedia Tools and Applications,2020,79(45):34587-34603.
[10] YUAN F,ZHANG L,WAN B,et al.Convolutional neural networks based on multi-scale additive merging layers for visual smoke recognition[J].Machine Vision and Applications,2019,30(2):345-358.
[11] YIN M,LANG C,LI Z,et al.Recurrent convolutional network for video-based smoke detection[J].Multimedia Tools and Applications,2019,78(1):237-256.
[12] XU G,ZHANG Q,LIU D,et al.Adversarial Adaptation From Synthesis to Reality in Fast Detector for Smoke Detection[J].IEEE Access,2019,7:29471-29483.
[13] TAO L,WANG X,YAMASAKI T.Rethinking Motion Representation:Residual Frames With 3D ConvNets[J].IEEE Transactions on Image Processing,2021,30:9231-9244.
[14] HE K,SUN J,TANG X.Single Image Haze Removal Using Dark Channel Prior[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(12):2341-2353.
[15] SANDLER M,HOWARD A,ZHU M,et al.Mobilenetv2:Inverted residuals and linear bottlenecks[C].proceedings of the Proceedings of the IEEE conference on computer vision and pattern recognition,F,2018.
[16] HOU Q,ZHOU D,FENG J.Coordinate attention for efficient mobile network design[C].proceedings of the Proceedings of the IEEE/CVF conference on computer vision and pattern recognition,F,2021.
[17] VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[J].Advances in neural information processing systems,2017:30.
[18] SRINIVAS A,LIN T Y,PARMAR N,et al.Bottleneck Transformers for Visual Recognition[C].proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR),2021.
[19] SHAW P,USZKOREIT J,VASWANI A.Self-Attention with Relative Position Representations[C].proceedings of the NAACL,F,2018.
[20] LI K,WANG Y,PENG G,et al.UniFormer: Unified Transformer for Efficient Spatial-Temporal Representation Learning[C].International Conference on Learning Representations,2022.
[21] LIU Z,MAO H,WU C-Y,et al.A convnet for the 2020s[C].proceedings of the Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,F,2022.
[22] KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM,2012,60:84-90.
[23] SIMONYAN K,ZISSERMAN A.Very Deep Convolutional Networks for Large-Scale Image Recognition[J].CoRR,2015.
[24] HE K,ZHANG X,REN S,et al.Deep Residual Learning for Image Recognition[C].proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR),2016.
[25] HOWARD A,SANDLER M,CHEN B,et al.Searching for MobileNetV3[C].proceedings of the 2019 IEEE/CVF International Conference on Computer Vision(ICCV),2019.
[26] LIU Z,LIN Y,CAO Y,et al. Swin Transformer:Hierarchical Vision Transformer using Shifted Windows[C].proceedings of the 2021 IEEE/CVF International Conference on Computer Vision(ICCV),2021.
[27] 张斌,魏维,高联欣,等.基于时空域深度神经网络的野火视频烟雾检测[J].计算机应用与软件,2019,36(9):236-242.
备注/Memo
收稿日期:2022-09-01
基金项目:四川省科技厅重点科研资助项目(2021YFG0299)