ResNet34

方法本研究收集了1420张舌象。经过筛选，专家对它们进行了判断，并对舌头图像进行了注释，以形成舌头图像数据集。然后将基于深度学习卷积神经网络 (CNN) 的人工智能网络框架，即ResNet34应用于该数据集，自动提取图像特征，实现舌图像分类。最后，将VGG16网络框架应用于数据集，对分类模型进行比较，并与分类效果进行比较。结果本文通过对标注采集的舌图像进行整理，形成了相关的数据集，验证了ResNet34架构能够更好地执行牙标和舌特征识别的任务

Methods: In this study, 1420 tongue images were collected. After screening, experts made judgments and annotated these tongue images to establish a tongue image dataset. Then, a deep learning convolutional neural network (CNN)-based artificial intelligence framework, namely ResNet34, was applied to this dataset to automatically extract image features and perform tongue image classification. Finally, the VGG16 network framework was applied to the dataset to compare the two classification models and their classification performance. Results: In this study, the collected annotated tongue images were organized to form a relevant dataset, and it was verified that the ResNet34 architecture can better perform the tasks of dental marker and tongue feature recognition.

搜集汇总

背景与挑战

背景概述

该数据集包含1420张经过专家标注的舌象图像，专为舌图像分类任务设计。研究采用ResNet34和VGG16等深度学习卷积神经网络框架进行特征提取和分类比较，验证了ResNet34在牙标和舌特征识别中的优越性能。数据集由泉州师范学院·厦门大学创新研究院于2022年发布，适用于医学图像分析和人工智能模型训练。