Calligraphy Digitization based on Node Sampling in Modeling of Cultural Communication Dynamic Network Span

Digital information processing is essential in the recent time. This paper studies the calligraphy digitization based on the node sampling in modeling of cultural communication dynamic network span. The design dynamic model uses the CPU and GPU combined model because the dynamic network pruning algorithm will not reduce the actual scale of the network model and there will be no obvious reasoning acceleration when running on the GPU, it is necessary to design a specific hardware accelerator for the dynamic pruning algorithm and process some steps in the hardware according to the process details of the algorithm. Thus, the deep neural network using dynamic pruning algorithm can realize obvious reasoning acceleration on the mobile terminal. Further, the node sampling algorithm is integrated to the calligraphy digitization process. In the experiment, the performance of the visualized result has been shown.