Tongue Image Quality Assessment Based on Deep Convolutional Neural Network

Abstract

Background: With the wide application of digital tongue diagnosis instrument, massive tongue images will be produced. Adequate image quality is the prerequisite to ensure accurate tongue image analysis. In the process of tongue image collection, improper operation may lead to many poor-quality images (fogging, underexposure, overexposure, blurred focus, wrong tongue posture, etc.), which seriously affect the image processing and the accuracy of image analysis. However traditional pattern recognition is difficult to evaluate the quality of tongue images by extracting features and manual removal of tongue images with bad quality consumes a lot of labor and has a high error rate. In this research, we utilized a deep convolutional neural network to automatically select bad quality tongue images.Methods: The present study was conducted to identify the most appropriate CNN model for Tongue Image Quality Assessment based on deep CNN. The CNN model was evaluated by using Residual neural network and compared with VGGNet and DenseNet. Evaluation metrics such as accuracy, precision, recall, and F1-score were used for CNN model performance.Results: A detection model is established for tongue image quality control based on deep residual network, with an average accuracy of 99.04%, accuracy of 99.05%, recall of 99.04%, and F1-score of 99.05%, which can be used for quality screening of massive tongue images.Conclusions: Our research findings demonstrate various CNN models in the decision-making process for the selection of tongue image quality assessment and prove that applying deep learning methods, specifically deep CNN, to evaluate bad quality tongue images is feasible.

Improved tongue detection algorithm based on YOLO V4

“…For example, Ling Zhang [6] et al proposed to use the characteristics of grayscale and color of the tongue image using a combination of grayscale histogram projection and adaptive thresholding segmentation method to achieve tongue segmentation. The algorithm uses grayscale projection of the tongue from different directions to determine the approximate area of the tongue, and then uses the OSTU method with adaptive thresholding to segment the tongue image, which only considers the grayscale features of the pixels and does not focus on other useful information of spatial characteristics.Li [7] et al used Convolutional Network (CNN) combined with HSV enhancement method to achieve tongue segmentation, which showed better results in the tongue segmentation task.Lin Bingqian [8] et al proposed a tongue segmentation scheme based on ResNet's deep convolutional neural network, which is trained by dividing the image into image blocks, detecting them based on the trained image blocks during recognition, and selecting the image block with the highest probability of containing the tongue in the image block for segmentation. Yiping Tang [9] et al proposed a multi-task convolutional neural network-based tongue image classification method that utilizes the correlation between labels to accomplish specific tasks at different network layers and uses multiple Softmax classifiers to achieve parallel prediction.…”

Development of attention-based robust deep learning model for tongue diagnosis by smartphone

“…There are methods of correcting by comparing pictures before and after using flash to improve accuracy and methods of filtering out low-quality tongue images by deep learning [31,32]. However, these methods cannot completely solve the influence of complex shooting conditions on tongue diagnosis.…”