Chinese character recognition has always attracted much attention in our country and is widely used in our lives and work. The combination of deep learning and neural networks and other algorithms has greatly improved the accuracy and speed. This article reviews the research status and related background of deep learning in the field of Chinese character recognition. First, it introduces the development process of Chinese character recognition and neural network algorithms. Secondly, it describes the Chinese character recognition architecture based on neural network, classifies and outlines the relevant methods of handwritten Chinese character recognition in simple scenarios and text recognition in complex scenarios, explains the construction and characteristics of network models, and analyzes and summarizes the characteristics of each network model. The shortcomings, and finally prospects for future research.