Dangerous Behavior Image Recognition Algorithm of Smart Port Based on Deep Neural Network

As an important manifestation of New Infrastructure, Smart port plays an important role in promoting the development of modern information technology and changing modern business model. However, the current Smart port staff have poor safety awareness, the port safety management is difficult to implement, the emergency response capacity is weak, and the awareness of dangerous behavior problems is not deep. This has led to frequent dangerous problems in the port, posing a serious threat to the safety of people's lives and property, and also greatly hindered the better development of the smart port. In order to better solve these problems, this paper conducts an in-depth study on the image recognition algorithm of dangerous behavior of Smart port. In this paper, a new deep learning-based image recognition method is proposed to address the problems of existing image recognition methods. Through the depth neural network, the Generative adversarial network model is improved, so that the generated images can be classified, and the image recognition algorithm is optimized. And to verify the effectiveness of DNN in the application of image discrimination algorithm for dangerous behaviors in smart ports, this paper compares it with the traditional image recognition algorithm. The research results show that when the number of dangerous behavior images of Smart port is 500, it has been experimentally verified that its recognition accuracy can reach 94.40%, and the time required for image recognition is 0.29 seconds. This shows that the algorithm in this paper has high recognition accuracy, fast recognition speed, and can effectively identify dangerous behaviors in the port, which plays an important role in ensuring the safety of life and property of Smart port staff.