Design and Implementation of the Chinese Character Font Recognition System Based on Binary Convolutional Encoding and Decoding Network

With the in-depth research of computer vision and natural language processing, Chinese character font recognition technology has been widely used. However, due to the large number, complex shape, and changeable style of Chinese characters, the traditional methods based on image feature extraction and classifier design are difficult to deal with the differences among different fonts effectively. Based on the above situation, this paper mainly studies designing and implementing the Chinese character font recognition system based on a binary convolutional encoding and decoding network (BCEDN). First, this paper introduces the principle and structure of Convolutional Encoding and Decoding Networks (CEDN) and how they can be applied to image generation tasks. Second, this paper proposes a binarization strategy to convert real-valued weights and activation functions in CEDN into binary weights and activation functions, thereby reducing the model's storage space and computational complexity while maintaining high generation quality. Then, the BCEDN model is trained and tested on this dataset. Finally, this paper analyses the performance of the BCEDN model on the Chinese character font recognition task through the results of quantitative and qualitative experiments and compares them with other related methods. The proposed BCEDN method has the advantages of high efficiency, stability, and scalability, having vital research significance in the field of Chinese character font recognition.