Mine external fire monitoring method using the fusion of visible visual features
-
摘要: 为了解决矿井外因火灾监测实时性差、误报率和漏报率高的问题,提出了基于可见光视觉特征融合的矿井外因火灾监测方法。首先,分析了不同监测环境中火源视频图像对应的视觉特征,并设计了火源纹理、尖角、相似系数和闪动频率的提取方法;其次,采用改进的种子区域生长算法对火灾疑似区域进行分割,并利用不同的火源特征提取方法计算火灾疑似区域的动、静态特征;然后,融合所提取的动、静态特征,构建火灾特征向量;最后,构建了基于BP神经网络的火灾监测模型,并对监测模型进行了验证。结果表明,本文提出的火灾监测方法可有效检测不同场景和不同距离下的外因火灾,正确率和检测率分别达到98.60%和99.09%,误检率低至2.00%,并有很强的抗干扰能力和鲁棒性。Abstract: In order to overcome the problems of poor real-time performance, high false alarm rate and underreport alarm rate of mine external fire monitoring, a method of fire monitoring using the fusion of visible visual features is proposed.Firstly, the visual features corresponding to the video images of fire sources in different monitoring environments are analyzed, and the extraction methods of fire source texture, sharp corners, similarity coefficient and flicker frequency are designed.Then, an improved seed region growth algorithm is used to segment the suspected fire area, and different feature extraction methods are used to calculate the dynamic and static characteristics of the suspected fire area.Secondly, the extracted dynamic and static features are used to construct fire feature vectors.Finally, a fire monitoring model using BP neural network is constructed, and monitoring model is verified.The results show that the proposed fire monitoring method can effectively detect mine external fire in different scenes and distances.The correct rate and detection rate are 98.60% and 99.06%, respectively, the false detection rate is as low as 2.00%.It has strong anti-interference ability and robustness.
-
Key words:
- external fire /
- fire monitoring /
- static features /
- dynamic features /
- feature fusion /
- BP neural networks
-
表 1 不同监控场景的视频数据信息
Table 1. Video data information of different monitoring scenarios
编号 目标类型 视频场景描述 视频/帧 1 火灾 早期点状火源,用于检测早期阶段的火灾 3 000 2 火灾 早期蔓延火源1,用于检测蔓延过程中的火灾 3 000 3 火灾 早期蔓延火源2,用于检测蔓延过程中的火灾 3 000 4 火灾 早期蔓延火源3,用于检测蔓延过程中的火灾 3 000 5 火灾 远距离点状火源,用于检测监视距离较远的火灾 3 000 6 干扰源 固定的巷道灯,用于排除井下固定的干扰光源 3 000 7 干扰源 矿工佩戴的矿灯,用于排除井下慢变的干扰光源 3 000 8 干扰源 设备的信号灯,用于排除井下闪烁的干扰光源 3 000 9 干扰源 正常行进的车灯,用于排除井下快变的干扰光源 3 000 10 干扰源 设备表面反射的光斑,用于排除井下的反射光源 3 000 表 2 不同算法的ACC指标值
Table 2. ACC values of different algorithms
算法 表 1中视频场景 1 2 3 4 5 6 7 8 9 10 Static & BP 55.26 81.39 59.26 68.94 51.54 94.42 48.84 89.21 92.05 93.14 Dynamic & BP 39.47 63.57 49.07 60.61 55.38 85.16 80.72 69.06 72.58 82.09 文献[22] 28.53 59.62 37.12 48.79 31.55 85.01 72.09 89.56 90.20 87.59 文献[23] 15.79 55.81 30.56 40.91 23.85 97.22 81.00 90.24 87.80 95.65 文献[10] 81.58 91.47 94.44 87.04 85.19 53.05 37.79 91.30 80.21 85.62 本文算法 57.89 71.67 65.37 62.88 42.31 99.01 82.78 92.81 93.42 97.86 -
[1] 孙继平, 孙雁宇, 范伟强. 基于可见光和红外图像的矿井外因火灾识别方法[J]. 工矿自动化, 2019, 45(5): 1-5, 21. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD201905001.htmSun Jiping, Sun Yanyu, Fan Weiqiang. Mine exogenous fire identification method based on visible light and infrared image[J]. Industry and Mine Automation, 2019, 45(5): 1-5, 21. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD201905001.htm [2] 孙继平, 范伟强. MS-ADoG域结合ReNLU与VGG-16的矿井双波段图像融合算法[J]. 光子学报, 2022, 51(3): 1-15. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB202203002.htmSun Jiping, Fan Weiqiang. Mine dual-band image fusion in MS-ADoG domain combined with ReNLU and VGG-16[J]. Acta Photonica Sinica, 2022, 51(3): 1-15. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB202203002.htm [3] 孙继平, 范伟强. 基于视频图像的瓦斯和煤尘爆炸感知报警及爆源判定方法[J]. 工矿自动化, 2020, 46(7): 1-4, 48. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202007001.htmSun Jiping, Fan Weiqiang. Gas and coal dust explosion perception alarm and explosion source judgment method based on video image[J]. Industry and Mine Automation, 2020, 46(7): 1-4, 48. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202007001.htm [4] 任慧, 孙继平, 刘妍. 用图像法探测煤矿井下火灾的研究[J]. 中国煤炭, 2005, 31(3): 59-60. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGME20050300L.htmRen Hui, Sun Jiping, Liu Yan. Research for the detection of mine fire by using graphical solution[J]. China Coal, 2005, 31(3): 59-60. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGME20050300L.htm [5] 王媛彬. 煤矿烟雾的计算机视觉检测方法[J]. 辽宁工程技术大学学报: 自然科学版, 2016, 35(11): 1230-1234. https://www.cnki.com.cn/Article/CJFDTOTAL-FXKY201611007.htmWang Yuanbin. Smoke detection based on computer vision in coal mine[J]. Journal of Liaoning Technical University: Natural Science, 2016, 35(11): 1230-1234. https://www.cnki.com.cn/Article/CJFDTOTAL-FXKY201611007.htm [6] 杜明本. 基于视频监控的煤矿井下火灾识别算法研究[D]. 太原: 太原科技大学, 2016. [7] 何晨阳, 周孟然, 闫鹏程, 等. 图像处理与BP神经网络在矿井火灾隐患识别中的应用[J]. 桂林理工大学学报, 2016, 36(3): 615-618. https://www.cnki.com.cn/Article/CJFDTOTAL-GLGX201603032.htmHe Chenyang, Zhou Mengran, Yan Pengcheng, et al. Application of image processing and BP neural network on mine fire potential hazard identification[J]. Journal of Guilin University of Technology, 2016, 36(3): 615-618. https://www.cnki.com.cn/Article/CJFDTOTAL-GLGX201603032.htm [8] 孙继平, 孙雁宇. 矿井火灾监测与趋势预测方法研究[J]. 工矿自动化, 2019, 45(3): 1-4. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD201903001.htmSun Jiping, Sun Yanyu. Research on methods of mine fire monitoring and trend prediction[J]. Industry and Mine Automation, 2019, 45(3): 1-4. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD201903001.htm [9] 王伟峰, 张宝宝, 王志强, 等. 基于YOLOv5的矿井火灾视频图像智能识别方法[J]. 工矿自动化, 2021, 47(9): 53-57. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202109008.htmWang Weifeng, Zhang Baobao, Wang Zhiqiang, et al. Intelligent identification method of mine fire video images based on YOLOv5[J]. Industry and Mine Automation, 2021, 47(9): 53-57. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202109008.htm [10] 孙继平, 崔佳伟. 矿井外因火灾感知方法[J]. 工矿自动化, 2021, 47(4): 1-5, 38. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202104001.htmSun Jiping, Cui Jiawei. Mine external fire sensing method[J]. Industry and Mine Automation, 2021, 47(4): 1-5, 38. https://www.cnki.com.cn/Article/CJFDTOTAL-MKZD202104001.htm [11] Deng L, Chen Q, He Y H, et al. Fire detection with infrared images using cascaded neural network[J]. Journal of Algorithms & Computational Technology, 2019, 13: 174830261989543. [12] 黄景雨. 林火虚警源红外成像特征建模及检测方法研究[D]. 成都: 电子科技大学, 2019. [13] Yang Z K, Wang T, Bu L P, et al. Training with augmented data: GAN-based flame-burning image synthesis for fire segmentation in warehouse[J]. Fire Technology, 2022, 58(1): 183-215. doi: 10.1007/s10694-021-01117-x [14] Ding H J, Gong F M, Gong W J, et al. Human activity recognition and location based on temporal analysis[J]. Journal of Engineering, 2018, 2018: 1-11. [15] 骆铁楠. 基于时间平滑多特征量火灾快速识别算法[J]. 煤炭技术, 2021, 40(5): 132-134. https://www.cnki.com.cn/Article/CJFDTOTAL-MTJS202105035.htmLuo Tienan. Fast recognition algorithm research of fire flame based on multifeatures longitude regression with temporal smoothing[J]. Coal Technology, 2021, 40(5): 132-134. https://www.cnki.com.cn/Article/CJFDTOTAL-MTJS202105035.htm [16] 王亚, 张宝峰. 基于显著性检测的红外森林火灾监测系统[J]. 消防科学与技术, 2018, 37(12): 1700-1703. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201812034.htmWang Ya, Zhang Baofeng. Infrared forest fire monitoring system based on saliency detection[J]. Fire Science and Technology, 2018, 37(12): 1700-1703. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201812034.htm [17] 范伟强, 刘毅. 基于自适应小波变换的煤矿降质图像模糊增强算法[J]. 煤炭学报, 2020, 45(12): 4248-4260. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202012030.htmFan Weiqiang, Liu Yi. Fuzzy enhancement algorithm of coal mine degradation image based on adaptive wavelet transform[J]. Journal of China Coal Society, 2020, 45(12): 4248-4260. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202012030.htm [18] 刘媛珺. 双波段野外火灾图像识别及目标定位方法研究[D]. 南京: 南京航空航天大学, 2009. [19] 辛颖, 李禹洁, 邢美净. 基于红外图像的森林凋落物阴燃火灾探测[J]. 消防科学与技术, 2018, 37(1): 61-64. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201801024.htmXin Ying, Li Yujie, Xing Meijing. Forest litter smoldering fire detection based on infrared image[J]. Fire Science and Technology, 2018, 37(1): 61-64. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201801024.htm [20] 杜志伟. 基于红外图像动态特征的早期火焰探测方法研究[D]. 天津: 天津大学, 2019. [21] 徐宏宇, 续婷. 一种基于颜色和纹理的优化SVM火灾识别方法[J]. 沈阳航空航天大学学报, 2021, 38(4): 54-60. https://www.cnki.com.cn/Article/CJFDTOTAL-HKGX202104012.htmXu Hongyu, Xu Ting. A color/texture-based improved SVM for fire recognition[J]. Journal of Shenyang Aerospace University, 2021, 38(4): 54-60. https://www.cnki.com.cn/Article/CJFDTOTAL-HKGX202104012.htm [22] 熊晓松, 刘钰如, 董逸君, 等. 基于机器视觉的火灾监测系统研究[J]. 消防科学与技术, 2017, 36(2): 217-220. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201702021.htmXiong Xiaosong, Liu Yuru, Dong Yijun, et al. Research on fire alarm monitor system based on machine vision[J]. Fire Science and Technology, 2017, 36(2): 217-220. https://www.cnki.com.cn/Article/CJFDTOTAL-XFKJ201702021.htm [23] 熊昊, 李伟. 基于SVM的视频火焰检测算法[J]. 传感器与微系统, 2020, 39(1): 143-145, 149. https://www.cnki.com.cn/Article/CJFDTOTAL-CGQJ202001040.htmXiong Hao, Li Wei. Video flame detection algorithm based on SVM[J]. Transducer and Microsystem Technologies, 2020, 39(1): 143-145, 149. https://www.cnki.com.cn/Article/CJFDTOTAL-CGQJ202001040.htm [24] 李晓宇, 陈伟, 杨维, 等. 基于超像素特征与SVM分类的人员安全帽分割方法[J]. 煤炭学报, 2021, 46(6): 2009-2022. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202106032.htmLi Xiaoyu, Chen Wei, Yang Wei, et al. Segmentation method for personnel safety helmet based on super-pixel features and SVM classification[J]. Journal of China Coal Society, 2021, 46(6): 2009-2022. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202106032.htm