Development and Analysis of Intelligent Pedestrian Prediction and Recognition System for Connected Vehicles

Intelligent networked prediction system is an emerging system developed in recent years, therefore, this paper is based on deep learning of the pedestrian recognition and behavior prediction system of intelligent networked vehicles, and its prediction effect is evaluated. Obtain the behavior information and characteristics of vehicles and pedestrians, construct a prediction database of intelligent vehicles and pedestrians, and adjust system parameters and indicators. Then, the information analysis of intelligent networked vehicles is realized, and the behavior content is analyzed with the help of deep learning, and the behavior data set is updated to complete the pedestrian recognition and behavior of intelligent networked vehicles. The simulation results show that deep learning can effectively improve the recognition rate of vehicle pedestrians and behaviors, increase by 5∼10%, and keep the recognition accuracy at 90%. At the same time, the prediction effect of vehicle behavior is improved, which can reach 85%, which is better than other methods. Compared with other methods, deep learning can predict vehicles and pedestrians, improve the overall system prediction effect, and provide support for the development of intelligent network prediction.