科技创新与应用2025，Vol.15Issue(34)：5-11,7.DOI:10.19981/j.CN23-1581/G3.2025.34.002
基于深度学习的药片识别技术研究进展
摘要
Abstract
Deep learning has been widely applied in the healthcare field,promoting the intelligent development of healthcare services.Currently,artificial intelligence mainly focuses on disease prediction in clinical environments,with relatively little research on the medications themselves.However,ensuring medication safety and avoiding medication errors is crucial for guaranteeing the correctness of the medications that patients take.This paper systematically outlines the relevant research and development history of pill detection and recognition technology,focusing on the methods based on deep learning.In deep learning,the application research of convolutional neural networks is particularly important.Compared with traditional machine learning methods,deep learning shows higher flexibility and accuracy in feature extraction.In addition,this paper also introduces public datasets related to tablet identification,and analyzes and prospects the problems faced by current tablet identification technology.关键词
深度学习/特征提取/卷积神经网络/目标检测/药片识别Key words
deep learning/feature extraction/convolutional neural network/target detection/pill recognition分类
信息技术与安全科学引用本文复制引用
王敏,欧翔,曹冉,潘发垒..基于深度学习的药片识别技术研究进展[J].科技创新与应用,2025,15(34):5-11,7.基金项目
国家自然科学基金资助项目(41775165,41775039) （41775165,41775039）
安徽省高校杰出青年科研项目(2023AH020022) （2023AH020022）
