采用机器学习的火焰前景提取算法

王博 1; 张为 1; 刘艳艳 2; 王鑫 3

doi:10.7652/xjtuxb201708005

您当前的位置：

首页 >

文章列表页 >

采用机器学习的火焰前景提取算法

更新时间：2025-07-09

- 采用机器学习的火焰前景提取算法
- An Extraction Algorithm of Fire Foreground Using Machine Learning Methods
- 西安交通大学学报 2017年51卷第8期页码：26-32+101
- 作者机构：
  
  天津大学电子信息工程学院,天津,300072
- 作者简介：
- 基金信息：
  
  : 2017-01-06。作者简介: 王博(1994—),男,硕士生;张为(通信作者),男,博士,教授。基金项目: 天津市科技支撑计划重点资助项目(14ZCZDSF00020)
- DOI：10.7652/xjtuxb201708005
  中图分类号： TP391.41
- 网络首发：2017-08-10，
  
  纸质出版：2017
- 稿件说明：
移动端阅览
王博 1, 张为 1, 刘艳艳 2, 等. 采用机器学习的火焰前景提取算法[J]. 西安交通大学学报, 2017,51(8):26-32+101.

An Extraction Algorithm of Fire Foreground Using Machine Learning Methods[J]. 2017, 51(8): 26-32+101.
王博 1, 张为 1, 刘艳艳 2, 等. 采用机器学习的火焰前景提取算法[J]. 西安交通大学学报, 2017,51(8):26-32+101. DOI： 10.7652/xjtuxb201708005.

An Extraction Algorithm of Fire Foreground Using Machine Learning Methods[J]. 2017, 51(8): 26-32+101. DOI： 10.7652/xjtuxb201708005.

摘要

针对现有火焰前景提取算法提取前景不完整、轮廓失真严重、对复杂环境适应性差等问题

提出一种采用机器学习的火焰前景提取算法。该算法结合使用监督学习方法和无监督学习方法

训练了两级专用的分类器用于确定疑似目标区域; 根据Real AdaBoost分类器的输出结果计算聚类算法的初始中心

并使用计算出的聚类中心对目标区域进行聚类分割

以得到最终的前景区域。实验结果表明:该算法对强光环境、夜间环境、静态或动态干扰环境等复杂场景均具有较好的适应性

得到的前景提取误差率在2%～28%之间

低于现有其他算法

且帧运算耗时小于50 ms

能够很好地完成多种场景下的火焰前景提取工作

为图像型火灾检测系统中后续的特征提取与识别奠定了基础。

Abstract

A novel extraction algorithm of fire foreground is proposed to solve the problems of incomplete foreground extraction

serious contour distortion and poor adaptability to complex environment in existing extraction algorithms. The algorithm combines supervised and unsupervised machine learning methods

and determine candidate fire regions by training a dedicated two-level classifier. An improved k-means algorithm is then adopted to segment the candidate regions

and initial cluster centers are calculated from the output of the Real AdaBoost classifier. Eventually

the clustered segmentation result is taken as the extractive foreground. Experimental results show that the proposed algorithm has good adaptability to various complex scenes

such as strong light environment

night lighting environment

and the environments with static or dynamic interference. The error rates of fire foreground extraction obtained by using the proposed algorithm range from 2% to 28%

that is

lower than those of other existing algorithms. Moreover

the processing time of each frame is shorter than 50 ms. The proposed algorithm can fulfill the fire foreground extraction task under various scenes and facilitate the succeeding feature extraction and recognition in image-type fire detection systems.

关键词

Keywords

references

KO B C, CHEONG K H, NAM J Y. Fire detection based on vision sensor and support vector machines [J]. Fire Safety Journal, 2009, 44(3): 322-329.

CHEN Juan, HE Yaping, WANG Jian. Multi-feature fusion based fast video flame detection [J]. Building Environment, 2010, 45(5): 1113-1122.

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.

WONG A K K, FONG N K. Experimental study of video fire detection and its applications [J]. Procedia Engineering, 2014, 71: 316-327.

ZHANG Haijun, ZHANG Nan, XIAO Nanfeng. Fire detection and identification method based on visual attention mechanism [J]. Optik-International Journal for Light and Electron Optics, 2015, 126(24): 5011-5018.

KOLESOV I, KARASEV P, TANNENBAUM A, et al. Fire and smoke detection in video with optimal mass transport based optical flow and neural networks [C]∥Proceedings of the 2010 17th IEEE International Conference on Image Processing. Piscataway, NJ, USA: IEEE, 2010: 761-764.

谢迪, 童若锋, 唐敏, 等. 具有高区分度的视频火焰检测方法 [J]. 浙江大学学报(工学版), 2012, 46(4): 698-704.

XIE Di, TONG Ruofeng, Tang Min, et al. Distinguishable method for video fire detection [J]. Journal of Zhejiang University(Engineering Science), 2012, 46(4): 698-704.

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.

TOREYIN B U, DEDEOGLU Y, CETIN A E. Flame detection in video using hidden Markov models [C]∥Proceedings of the 2005 International Conference on Image Processing. Piscataway, NJ, USA: IEEE, 2006: 1230-1233.

SCHAPIRE R E, SINGER Y. Improved boosting algorithms using confidence-rated predictions [J]. Machine Learning, 1999, 37(3): 297-336.

张进华, 李婷, 王孙安, 等. 可变视场下的火灾探测算法 [J]. 西安交通大学学报, 2012, 46(10): 29-35.

ZHANG Jinhua, LI Ting, WANG Sun'an, et al. An early detection algorithm for small flame based on vision system with variant field of view [J]. Journal of Xi'an Jiaotong University, 2012, 46(10): 29-35.

耿庆田, 于繁华, 赵宏伟, 等. 基于颜色特征的火焰检测新算法 [J]. 吉林大学学报(工学版), 2014, 44(6): 1787-1792.

GENG Qingtian, YU Fanhua, ZHAO Hongwei, et al. New algorithm of flame detection based on color features [J]. Journal of Jilin University(Engineering and Technology Edition), 2014, 44(6): 1787-1792.

CETIN A E, DIMITROPOULOS K, GOUVERNEUR B, et al. Video fire detection: review [J]. Digital Signal Processing, 2013, 23(6): 1827-1843.

TRAMBITCKII K, ANDING K, MUSALIMOV V, et al. Colour based fire detection method with temporal intensity variation filtration [C]∥Proceedings of the 2014 Joint IMEKO TC1-TC7-TC13 Symposium: Measurement Science Behind Safety and Security. Bristol, UK: IOP Publishing, 2015, 588(1): 012038.

TANG Yunwei, JING Linhai, Li Hui, et al. A multiple-point spatially weighted k-NN method for object-based classification [J]. International Journal of Applied Earth Observation Geoinformation, 2016, 52: 263-274.

LAI Yingxun, LAI Chinfeng, HUANG Yuehmin, et al. Multi-appliance recognition system with hybrid SVM/GMM classifier in ubiquitous smart home [J]. Information Sciences, 2013, 230(4): 39-55.

RUTKOWSKI L, JAWORSKI M, PIETRUCZUK L, et al. The CART decision tree for mining data streams [J]. Information Sciences, 2014, 266(5): 1-15.

OJALA T, PIETIKAINEN M, MAENPAA T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns [J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2002, 24(7): 971-987.

VIOLA P, JONES M. Fast and robust classification using asymmetric AdaBoost and a detector cascade [J]. Advances in Neural Information Processing Systems, 2002, 14: 1311-1318.

KO B C, CHEONG K H, NAM J Y. Early fire detection algorithm based on irregular patterns of flames and hierarchical Bayesian networks [J]. Fire Safety Journal, 2010, 45(4): 262-270.

LI H, NGAN K N. Unsupervised video segmentation with low depth of field [J]. IEEE Transactions on Circuits Systems for Video Technology, 2008, 17(12): 1742-1751.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

机器学习在沸腾传热过程中的应用进展

跨声速压气机转子通用k-ω湍流模型的机器学习辅助优化

面向CO2捕集的吸附剂研究：从传统设计到机器学习

一种基于边界识别的聚类算法

电动车辆间接热系统快速除湿响应迟滞机理及优化控制策略