Real-Time Fire Detection Based on Deep Convolutional Long-Recurrent Networks and Optical Flow Method

A new fire monitoring method was proposed in this paper, which proposed a Neural Network of Deep Convolutional Long-Recurrent Networks (DCLRN), and combining DCLRN network and optical flow method for fire monitoring in open space environment in real time. This is achieved by utilizing the static and dynamic characteristics of the fire, converting fire RGB images to optical flow images in real-time, and use convolutional neural network for spatial learning, a class of recurrent convolutional architectures for sequence learning, eventually achieve the purpose of fire monitoring. Which is end-to-end trainable and suitable for large-scale visual understanding tasks of fire monitoring. Our novelties are: firstly, our method is the first to our knowledge to put forward DCLRN, and combining DCLRN with optical flow method for fire monitoring. Secondly, our method has the ability that can detect the smoke as well as the flames. Finally, in this way the fire can be detected as soon as it occurs, achieved early detection of fire. The experiments have proved that DCLRN combined with optical flow images have good accuracy and reliability in the detection and recognition of fire monitoring videos, and give good performance on a more challenging dataset.