J Shanghai Jiaotong Univ Sci

Journal of Shanghai Jiaotong University(Science) >

2025 , Vol. 30 >Issue 5: 866 - 879

DOI: https://doi.org/10.1007/s12204-023-2680-1

Computing & Computer Technologies

Multi-Scene Smoke Detection Based on Multi-Feature Extraction Method

Expand
  • 1. School of Computer Science, Hangzhou Dianzi University, Hangzhou 310018, China; 2. Deep Space Exploration Lab, Beijing 100080, China

Received date: 2023-06-29

  Accepted date: 2023-07-20

  Online published: 2023-12-21

Fold

Abstract

This study proposes a multi-scene smoke detection algorithm based on a multi-feature extraction method to address the problems of varying smoke shapes in different scenes, difficulty in locating and detecting translucent smoke, and variable smoke scales. First, the convolution module of feature extraction in YOLOv5s backbone network is replaced with asymmetric convolution block re-parameterization convolution to improve the detection of different shapes of smoke. Then, coordinate attention mechanism is introduced in the deeper layer of the backbone network to further improve the localization of translucent smoke. Finally, the detection of smoke at different scales is further improved by using the feature pyramid convolution module instead of the standard convolution module of the feature pyramid in the model. The experimental results demonstrate the feasibility and superiority of the proposed model for multi-scene smoke detection.

Cite this article

SHAO Yanli, YING Yong, CHEN Xi, DONG Siyu, WEI Dan . Multi-Scene Smoke Detection Based on Multi-Feature Extraction Method[J]. Journal of Shanghai Jiaotong University(Science), 2025 , 30(5) : 866 -879 . DOI: 10.1007/s12204-023-2680-1

References

[1] JIANG T Y, LI C, YANG M, et al. An improved YOLOv5s algorithm for object detection with an  attention mechanism [J]. Electronics, 2022, 11(16): 2494. 

[2] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection  and semantic segmentation [C]//2014 IEEE Conference  on Computer Vision and Pattern Recognition. Columbus: IEEE, 2014: 580-587. 

[3] DING H D, LIU S Y, XU Y K, et al. Novel video smoke  observation and detection method based on the motion  track block [J]. Journal of Safety and Environment, 2016, 16(4): 96-100 (in Chinese). 

[4] YE S P, BAI Z C, CHEN H F, et al. An effective algorithm  to detect both smoke and flame using color  and wavelet analysis [J]. Pattern Recognition and Image Analysis, 2017, 27(1): 131-138. 

[5] ZHAO L, LUO Y M, LUO X Y. Based on dynamic  background update and dark channel prior of fire  smoke detection algorithm [J]. Application Research of Computers, 2017, 34(3): 957-960 (in Chinese). 

[6] WANG S D, HE Y P, YANG H Y, et al. Video smoke  detection using shape, colorand dynamic features [J]. Journal of Intelligent & Fuzzy Systems: Applications  in Engineering and Technology, 2017, 33(1): 305-313. 

[7] 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 [J]. Fire Safety Journal, 2018, 102: 1-10. 

[8] LUO Y M, ZHAO L, LIU P Z, et al. Fire smoke detection  algorithm based on motion characteristic and convolutional  neural networks [J]. Multimedia Tools and Applications, 2018, 77(12): 15075-15092. 

[9] LI T T, ZHAO E T, ZHANG J G, et al. Detection of  wildfire smoke images based on a densely dilated convolutional  network [J]. Electronics, 2019, 8(10): 1131. 

[10] XU G, ZHANG Y M, ZHANG Q X, et al. Video  smoke detection based on deep saliency network [J]. Fire Safety Journal, 2019, 105: 277-285.

[11] MUHAMMAD K, KHAN S, PALADE V, et al. Edge  intelligence-assisted smoke detection in foggy surveillance  environments [J]. IEEE Transactions on Industrial Informatics, 2020, 16(2): 1067-1075. 

[12] LI C H, YANG B, DING H, et al. Real-time videobased  smoke detection with high accuracy and efficiency [J]. Fire Safety Journal, 2020, 117: 103184. 

[13] HE L J, GONG X L, ZHANG S R, et al. Efficient attention  based deep fusion CNN for smoke detection in  fog environment [J]. Neurocomputing, 2021, 434: 224- 238. 

[14] ZHANG F, QIN W, LIU Y B, et al. A dual-channel  convolution neural network for image smoke detection [J]. Multimedia Tools and Applications, 2020, 79(45/46): 34587-34603. 

[15] LIU S, QI L, QIN H F, et al. Path aggregation network  for instance segmentation [C]//2018 IEEE/CVF Conference  on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8759-8768. 

[16] DING X H, GUO Y C, DING G G, et al. ACNet: Strengthening the kernel skeletons for powerful CNN via asymmetric convolution blocks [C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 1911-1920. 

[17] RAMACHANDRAN P, ZOPH B, LE Q V. Searching  for activation functions [DB/OL]. (2017-10-16) [2023- 06-29]. https://arxiv.org/abs/1710.05941 

[18] ZEILER M D, FERGUS R. Visualizing and understanding  convolutional networks [M]//Computer  vision–ECCV 2014. Cham: Springer International Publishing, 2014: 818-833. 

[19] HOU Q B, ZHOU D Q, FENG J S. Coordinate attention  for efficient mobile network design [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 13708-13717. 

[20] DUTA I C, LIU L, ZHU F, et al. Pyramidal convolution: Rethinking convolutional neural networks for  visual recognition [DB/OL]. (2020-06-20) [2023-06-29]. https://arxiv.org/abs/2006.11538 

[21] REN S Q, HE K M, GIRSHICK R, et al. Faster RCNN: Towards real-time object detection with region  proposal networks [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137- 1149. 

[22] LIUW, ANGUELOV D, ERHAN D, et al. SSD: Single  shot MultiBox detector [M]//Computer vision–ECCV 2016. Cham: Springer International Publishing, 2016: 21-37. 

[23] GE Z, LIU S T, WANG F, et al. YOLOX: Exceeding YOLO series in 2021 [DB/OL]. (2021-07-18) [2023-06- 29]. https://arxiv.org/abs/2107.08430 

[24] SHI H. Research on field fire smoke detection  method based on deep neural network [D]. Hangzhou: Hangzhou Dianzi University, 2021 (in Chinese). 

[25] JADON A, OMAMA M, VARSHNEY A, et  al. FireNet: A specialized lightweight fire &  smoke detection model for real-time IoT applications [DB/OL]. (2019-05-28) [2023-06-29]. https://arxiv.org/abs/1905.11922 

[26] ZHENG Y P, XU B Y, WANG Z Y. Improved YOLOv5 smoke detection model [J]. Computer Engineering  and Applications, 2023, 59(7): 214-221 (in Chinese).

Options
Outlines

/

〈 〉