Red, White, Crimson and Purple Tongue Diagnosis Based on Deep Learning

Tongue diagnosis in traditional Chinese medicine is a very important diagnostic technique, so now the research on Objectification of tongue diagnosis is becoming more and more important, and in-depth learning is just the research focus in the field of tongue diagnosis. In this study, we use convolutional neural network to analyze the tongue diagnosis pictures. First, we collect the tongue diagnosis pictures of traditional Chinese medicine on the Internet, modify and screen them, establish a data set image required for this experiment, and randomly divide the data set into training set and test set for the experimental research of this paper. The division ratio is 4:1. Then the convolutional neural network model (CNN) is built to train and test the data. Then we use KNN and SVM models to conduct comparative experiments, and build a table of the experimental results for the final result analysis.