1. Institute of Electrical Engineering, Yanshan University, Qinhuangdao, Hebei 066004, China
2. Yanshan University Science Park, Qinhuangdao, Hebei 066004, China
Abstract：A smoke detection algorithm based on improved Gaussian mixture and YOLOv2 is proposed to detect fire timely and effectively. First of all, according to the characteristics of the early smoke makes the improvement to the Gaussian mixture, effectively framing the dynamic target region of interest, to extract the smoke foreground based on smoke detection; converted to regression problems, using end-to-end target detection algorithm YOLOv2 smoke training data set, second detection and screening, the final box set and the specific location the scope of the smoke area, meet the effective detection of the different scenes of smoke. The experimental results show that the fusion algorithm improves the detection effect of smoke area, improves the accuracy and effectively reduces the false detection rate to the smoke.
［1］Mokhtari Z, Holé S, Lewiner J. Temperature and humidity behaviour of a corona based smoke sensor ［J］. Journal of Electrostatics, 2015, 76: 62-66.
［2］Yang S L, Jin S. Analysis of the Optical Smoke Detector Sensitivity ［J］. Chinese Journal of Sensors & Actuators, 2006, 19(2): 425-428.
［3］Ko B C, Park J O, Nam J Y. Spatiotemporal bag-of-features for early wildfire smoke detection ［J］. Image and Vision Computing, 2013, 31(10): 786-795.
［4］Simone Calderara, Paolo Piccinini, Rita Cucchiara. Vision based smoke detection system using image energy and color information ［J］ . Machine Vision and Applications, 2011, 22(4): 705-719.
［5］Morerio P, Marcenaro L, Regazzoni C S, et al. Early fire and smoke detection based on colour features and motion analysis ［C］// Image Processing (ICIP), 2012 19th IEEE International Conference on. IEEE, 2012: 1041-1044.
［6］Marko Heikkil, Matti Pietikinen, Schmid C. Description of interest regions with local binary patterns ［J］. Pattern recognition, 2009, 42(3): 425-436.
［7］程淑红, 高许, 程树春, 等. 基于计算机视觉的运动车辆检测［J］. 计量学报, 2017, 38(3): 288-291.
Cheng S H, Gao X, Cheng S C, et al. Moving Vehicle Detection Based on Computer Vision ［J］. Acta Metrologica Sinica, 2017, 38(3): 288-291.
［8］Viola P, Jones M. Rapid object detection using a boosted cascade of simple features ［C］// Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on. 2001, 1: I-I.
［9］Stauffer C, Grimson W E L. Adaptive background mixture models for real-time tracking ［C］// Computer Vision and Pattern Recognition, 1999. IEEE Computer Society Conference on. 1999, 2: 246-252.
［10］James M. Flame and Smoke Estimation for Fire Detection in Videos based on Optical Flow and Neural Networks ［J］. International Journal of Research in Engineering and Technology, 2014, 3(8): 324-328.
［11］王洪斌, 于菲, 李一骏, 等. 分块特征匹配与局部差分结合的运动目标检测［J］. 计量学报, 2015, 36(4): 352-355.
Wang H B, Yu F, LI Y J, et al. Detection of Moving Object by Combining Block Features Matching and Local Differential ［J］. Acta Metrologica Sinica, 2015, 36(4): 352-355.
［12］Redmon J, Farhadi A. YOLO9000: better, faster, stronger ［C/OL］.https://arxiv.org/abs/1612.08242.
［13］Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection ［C］//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016: 779-788.
［14］Cooijmans T, Ballas N, Laurent C, et al. Recurrent batch normalization ［C/OL］. https://arxiv.org/abs/1603.09025.
［15］Ren S, He K, Girshick R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks ［C］//Advances in neural information processing systems. Montreal, Canada, 2015: 91-99.
2019, Vol. 40 
