Journal of Shandong University(Engineering Science) ›› 2019, Vol. 49 ›› Issue (1): 1-9.doi: 10.6040/j.issn.1672-3961.0.2017.430

• Machine Learning & Data Mining •     Next Articles

Video flame detection based on GMM and 3D-LBP feature

Yunyang YAN1,2(),Huizhen ZHANG1,2,Yi′an LIU2,Shangbing GAO1   

  1. 1. Faculty of Computer & Software Engineering, Huaiyin Institute of Technology, Huaian 223003, Jiangsu, China
    2. School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, Jiangsu, China
  • Received:2017-08-29 Online:2019-02-20 Published:2019-03-01
  • Supported by:
    国家自然科学基金(61402192);江苏省“六大人才高峰”(2013DZXX-023);江苏省“333工程”(BRA2013208);江苏省“青蓝工程”(2017);淮安市“533工程”(2017);淮安市科技计划(HAG2013057)

RichHTML

175

PDF (PC)

540

Abstract:

In order to solve the problems of extracting the accuracy of the candidate region and improve the description ability of the flame characteristics, a novel flame detection algorithm based on Gaussian mixture model (GMM) and three-dimensional locality binary pattern (LBP) texture features was proposed. The distribution of flame was analyzed in two spaces of RGB and HSV, and the GMM was trained to extract the flame candidate region. The texture characteristics of the flame was selected as an important feature. The original LBP texture was fused with the motion characteristics of the flame to form a new three-dimensional LBP texture to improve the classification effect of the texture feature on the flame. The one-class support vector machine (One-Class SVM) classification method was used to determine whether the candidate area was a flame.

Key words: flame detection, GMM, dynamic feature, 3D-LBP, SVM

CLC Number: 

  • TP391

Fig.1

Extraction of candidate regional based on the Gaussian mixture model of motion and color"

Fig.2

Extraction process of 2D LBP texture"

Fig.3

Space map of 3D dynamic LBP texture feature extraction"

Fig.4

Invariance of plane rotation of 3D LBP texture"

Fig.5

Merging histograms of each component of flameRGB color space"

Fig.6

Classification charts of One-class SVM"

Table 1

Comparison of flame detection accuracy with different kernel functions"

核函数 准确检测帧数 检测率/%
sigmoid核 577 96.1
多项式核 587 97.8
高斯核 589 98.2

Fig.7

Flame region after the discrimination by SVM"

Fig.8

Flow chart of algorithm"

Table 2

Description of database video sets"

视频编号 视频内容描述
1 庭院火焰,鼎内火焰透明度很高,飘忽不定,墙壁与火焰颜色相似
2 草地火焰,草丛中有人走动,火焰面积由小到大
3 森林火焰,图像中心有大面积火焰,色彩较鲜艳
4 庭院火焰,图像整体偏暗,有小面积火焰,初时火焰颜色不明显,有人在其旁边经过
5 森林火焰,有较多中小面积火焰区域,上空伴有浓烟
6 荒野火焰,土壤颜色与火焰类似,右下角火焰不突出
7 公路场景,在夜晚中一辆汽车从远处驶来,白色强光照亮地面
8 公路场景,夜色下一辆卡车经过,地面上有强烈的白色灯光伴黄色光晕,路边有行人,四周有其它车灯交杂

Fig.9

Detection of results sample videos"

Table 3

Results of flame detection video"

视频 总帧数 火焰帧数 正检火焰帧 文献[13] 文献[14] 文献[15] 本研究算法
TP/% FP/% TP/% FP/% TP/% FP/% TP/% FP/%
1 429 429 425 89.4 10.6 93.9 6.1 95.9 4.1 99.1 0.9
2 250 250 250 95.7 4.3 95.1 4.9 97.6 2.4 100 0
3 236 236 233 84.7 15.3 96.2 3.8 95.8 4.2 98.7 1.3
4 320 320 275 73.3 26.7 83.0 17.0 92.1 7.9 85.9 14.1
5 235 235 234 84.3 15.7 95.9 4.1 94.9 5.1 99.6 0.4
6 290 290 451 90.7 9.3 93.2 6.8 91.7 8.3 98.9 1.1

Table 4

Results of non flame video detection"

视频 总帧数 火焰帧数 误检帧数 文献[13] 文献[14] 文献[15] 本研究算法
TN/% FN/% TN/% FN/% TN/% FN/% TN/% FN/%
7 146 0 0 46.7 53.3 91.4 8.3 92.0 8.0 100 0
8 151 0 2 22.8 56.5 87.5 12.5 96.4 3.6 98.7 1.3

Table 5

Detection time of algorithm"

算法 文献[13] 文献[14] 文献[15] 本研究算法
平均每帧检测时间/s 1.112 0.783 0.946 0.453
1 CHI R , LU Z M , JI Q G . Real-time multi-feature based fire flame detection in video[J]. IET Image Processing, 2017, 11 (1): 31- 37.
2 KONG S G , JIN Donglin , LI Shengzhe , et al. Fast fire flame detection in surveillance video using logistic regression and temporal smoothing[J]. Fire Safety Journal, 2016, 79, 37- 43.
3 JIANG Bo , LU Yongyi , LI Xiying , et al. Towards a solid solution of real-time fire and flame detection[J]. Multimedia Tools and Applications, 2015, 74 (3): 689- 705.
4 吴建胜, 高云玲, 张宾. 一种基于视频的火焰检测方法[J]. 计算机应用与软件, 2014, 31 (1): 173- 175, 190.
WU Jiansheng , GAO Yunling , ZHANG Bin . A video-based flame detection method[J]. Computer Applications and Software, 2014, 31 (1): 173- 175, 190.
5 DIMITROPOULOS K , BARMPOUTIS P , GRAMMALIDIS N . Spatio-temporal flame modeling and dynamic texture analysis for automatic video-based fire detection[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2015, 25 (2): 339- 351.
6 GANESAN P, RAJINI V, SATHISH B S, et al. HSV color space based segmentation of region of interest in satellite images[C]//Proceedings of the 2014 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT). Kanyakumari, India: IEEE, 2014, 12(10): 101-105.
7 乔少杰, 金琨, 韩楠, 等. 一种基于高斯混合模型的轨迹预测算法[J]. 软件学报, 2015, 26 (5): 1048- 1063.
QIAO Shaojie , JIN Kun , HAN Nan , et al. Trajectory prediction algorithm based on Gaussian mixture model[J]. Journal of Software, 2015, 26 (5): 1048- 1063.
8 唐岩岩, 严云洋, 刘以安. 应用GMM的快速火焰检测[J]. 计算机科学, 2012, 39 (11): 283- 285.
TANG Yanyan , YAN Yunyang , LIU Yian . Fast flame detection with GMM[J]. Computer Science, 2012, 39 (11): 283- 285.
9 DOSHI N P, SCHAEFER G, HOSSAIN S. Rotation invariant compound LBP texture features[C]//Proceedings of the 5th International Conference on Informatics, Electronics and Vision (ICIEV). Dhaka, Bangladesh: IEEE, 2016: 1057-1061.
10 KUMAR A R , SARAVANAN D . Content based image retrieval using color histogram[J]. Int. J. Comput. Sci. Inform. Technol, 2013, 4, 242- 245.
11 王洪波.单分类支持向量机的学习方法研究[D].苏州:浙江大学, 2012.
WANG Hongbo. Research of learning method of single-class support vector machine[D]. Suzhou: Zhejiang University, 2012.
12 GU B , SHENG V S , TAY K Y , et al. Incremental support vector learning for ordinal regression[J]. IEEE Transactions on Neural Networks and Learning Systems, 2015, 26 (7): 1403- 1416.
13 CHEN T H, WU P H, CHIOU Y C. An early fire-detection method based on image processing[C]//Proceedings of the 2004 International Conference on Image Processing. Singapore City, Singapore: IEEE, 2004, 3: 1707-1710.
14 TUNG X T , 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.
15 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.
[1] LIU Yan, LI Youjun, CHEN Meng. Research on the classification of resting state EEG signal between depression patients and normal controls by EMD and SVM methods [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(3): 21-26.
[2] ZHANG Yuling, YIN Chuanhuan. Android malware detection based on SVM [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(1): 42-47.
[3] YAN Yun-yang1,2, TANG Yan-yan2, LIU Yi-an2, ZHANG Tian-yi3. Flame detection based on LBP features with multiscales and SVM [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2012, 42(5): 47-52.
[4] PAN Dong-yin, ZHU Fa, XU Sheng, YE Ning*. Feature selection of gene expression profiles of colon cancer [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2012, 42(2): 23-29.
[5] LIU Hai-chun,XIE Shao-jun,WANG Guo-feng . A novel hybrid current control method applied in threephase four leg APF [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2007, 37(6): 10-14 .
[6] WANG Sheng-chun,HAN Jie,LI Zhi-nong,LI Jian-feng . The fault diagnosis method based on ambiguity domain and support vector machine [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2006, 36(6): 116-120 .
[7] ZHU Xiao-xia,SUN Tong-jing,CHEN Gui-you . Research of fingerprint classification combined by binary tree and SVM [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2006, 36(1): 121-124 .
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
[1] XIA Bin,ZHANG Lian-jun . Energy comparison-based TOA estimation algorithm for the DS-CDMA UWB system[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2007, 37(1): 70 -73 .
[2] BO De-Yun, ZHANG Dao-Jiang. Adaptive spectral clustering algorithm[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2009, 39(5): 22 -26 .
[3] ZHAO Ke-Jun, WANG Xin-Jun, LIU Xiang, CHOU Yi-Hong. Algorithms of continuous top-k join query over structured overlay networks[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2009, 39(5): 32 -37 .
[4] DING Wan-Tao, LI Shu-Cai, ZHANG Qing-Song. Discussion on interface error regularity of inclined  stratum predicted by TSP[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2009, 39(4): 57 -60 .
[5] WANG Bai-wei,CAO Sheng-le . A mult-objective assessment method of the effects of industrial waste-water management[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2007, 37(3): 89 -92 .
[6] CHOU Wu-Sheng, WANG Shuo. Study on the adaptive algorithm of the force reflection robotic master under large stiffness of the environment[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2010, 40(1): 1 -5 .
[7] ZHANG Hui，WANG Meng-xia, HAN Xue-shan. The advanced thermal rating of power system and its application[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2008, 38(6): 25 -29 .
[8] YAN Chong-jing, LIAO Wen-he, GUO Yu, CHENG Xiao-sheng. The BOM modeling based on the polychromatic graph[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2008, 38(6): 70 -75 .
[9] WANG Jian-ping,WANG Shu-hua,GENG Gui-li . Study on the transformation activation energy in InN semiconductor nanocrystals[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2008, 38(2): 42 -44 .
[10] CHEN Bin, CHEN Song-Can, PAN Zhi-Song, LI Bin. Survey of outlier detection technologies[J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2009, 39(6): 13 -23 .
Copyright 2018 © Editorial office of Journal of Shandong University (Engineering Science)
Supported by Beijing Magtech Co.Ltd