Research on Fire Detection Method of Surveillance Video in Laboratory Management Based on Deep Learning and Motion Features

Fire detection of surveillance video in laboratory management is an important security work. In this paper, a fire detection method based on deep learning and motion features is proposed to improve the accuracy and real-time performance of fire recognition. In the first part, this paper introduces the application background of fire surveillance video and the application of deep learning in video analysis. A fire detection system, including a data acquisition module, video preprocessing module, and fire recognition module, is designed. In the fire recognition module, we adopted a hybrid model based on a convolutional neural network and a long short-term memory network. We used motion features and visual features to judge the situation. Furthermore, we trained and optimized the model and verified the effectiveness and performance of the model through experiments. The experimental results show that the method proposed in this paper has high accuracy, recall ratio, and F1 value in fire recognition and can realize real-time processing. This paper proposes a new scheme that provides new ideas and technologies for fire detection and video surveillance in laboratory management.