Study on Smoking and Telephone Behaviour Detection Based on Convolutional Neural Network

Abstract

Smoking and making phone calls are common behaviors in life, but in some cases, these two behaviors are inappropriate and even bring serious harm to people and the environment. At present, these two methods are mainly detected by smoke alarm and signal detector. This paper mainly introduces how to judge these two behaviors by connecting the camera to the computer through computer vision and deep learning. The algorithm used in this paper is convolutional neural network. The content includes the process of data set processing, the introduction of the concept of convolutional neural network, the role of each layer of the model in this paper, and how the algorithm works. In addition, this paper connects the trained convolutional neural network model with the camera to detect people’s behavior in the camera in real time. This study first converts the images in the data set into many matrices of a certain size after a series of processing and divided these matrices into test set and training set. The video in the camera is converted into a picture stream and input into the trained model to achieve the effect of real-time detection. The results show that this algorithm has good accuracy. It is feasible to detect people’s smoking and telephone behavior in this way.