Research On Smoke Detection Algorithm Based On Deep Learning

Fire is one of common and highly hazardous disasters.Smoke detection is an effective means of early fire alarming,due to the smoke,which produced by the combustion of materials,spreads significantly faster than the flame.There are two branches of smoke detection,one of which is smoke detection based on contact sensors,and the other one is smoke detection based on image.Among them,smoke detection based on contact sensors has been matured after decades of development,but it has the inherent disadvantages of small detection range and high installation cost.Image-based smoke detection technology can solve these problems,which is the focus of the development of early fire warnings in recent years.It is a challenging task to recognize smoke from images due to large variance of smoke color,texture and shapes.There are smoke detection methods that have been proposed,but most of them are based on hand-crafted features.To improve the performance of smoke detection,two research objects have been completed by us,we propose a novel deep normalization and convolutional neural network(DNCNN)with fourteen layers to implement automatic feature extraction and classification.In DNCNN,traditional convolutional layers are replaced with normalization and convolutional layers to accelerate the training process and boost the performance of smoke detection.To reduce overfitting caused by imbalanced and insufficient training samples,we generate more training samples from original training data sets by using a variety of data enhancement techniques.Experimental results show that our method achieved very low false alarm rates below 0.60% with detection rates above 96.37% on our smoke datasets.