National Cultural Symbols Recognition Based on Convolutional Neural Network

In order to solve the problem that the process of manually identifying national symbols is extremely tedious and the recognition effect is not satisfactory, the paper uses the TensorFlow framework to build a convolutional neural network to identify domestic symbols simply and efficiently. In this paper, the classified Zhuang ethnic symbol pictures are labeled and normalized to make a data set, and then during the training process, the loss value between the prediction result and the correct answer is continuously reduced to train a convolution layer, pool The convolutional neural network of the visualization layer, the fully connected layer, and the SoftMax layer. Finally, the images are classified by the SoftMax layer. The experimental results show that after a lot of training, the model has been more robust, and the recognition rate of 15 symbol types can reach 89%, which is faster and more accurate than the manual recognition process.