A method and system for determining a tongue picture sample library based on a deep learning model

Abstract

The invention provides a method and system for determining a tongue picture sample library based on a deep learning model. The method and the system are characterized by marking each tongue image feature of all tongue sample images in a tongue image sample set, establishing a deep learning model after labeling; determining the accuracy of tongue image feature labeling through training and testing;when the accuracy does not reach the set accuracy, revising the tongue image characteristics in the tongue image sample set on the basis of the original annotation; and determining the marking accuracy through the deep learning model again, performing multi-round marking on each tongue image feature, and detecting the marking accuracy through the multi-round deep learning model to complete the determination of the marking accuracy of all tongue image features, thereby determining the tongue image sample library of the traditional Chinese medicine tongue diagnosis. According to the method andthe system, the tongue image characteristics are subjected to multi-round labeling and multi-time deep learning model training to detect the accuracy of the tongue image characteristics, so that the tongue image labeling accuracy is greatly improved.

Description

technical field [0001] The present invention relates to the field of medical image calibration, and more specifically, to a method and system for determining a tongue image sample library based on a deep learning model. Background technique [0002] At present, the calibration technology for medical images is not very mature, especially in the calibration of TCM tongue images. Due to the similarity of tongue image color, the complexity of tongue shape, and the different characteristics of each tongue image, many calibration methods cannot be directly used in the calibration of TCM tongue image, and the algorithm has high requirements for the accuracy of labeling. . [0003] For the current image annotation technology in other fields, it mainly uses traditional computer vision algorithms for automatic annotation, such as image feature extraction algorithms, which extract the features of the image to locate the key positions in the image for annotation. However, the traditio...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the technical problem of low accuracy of tongue image feature labeling of samples in the tongue image sample library in the prior art, the present invention provides a method for determining the tongue image sample library based on a deep learning model, the method comprising:

Images