深度学习在数字图像档案智能检索中的应用研究  

The Research of Digital Images Archives Based on Deep Learning

作　　者：李娟[1] 王瑞良 LI Juan;WANG Rui-liang(Zhongyuan University of Technology,Zhengzhou Henan 450007;Zhengzhou Branch of China Radio and Television Henan Network Co.,Ltd.,Zhengzhou Henan 450000)

机构地区：[1]中原工学院,河南郑州450007 [2]中国广电河南网络有限公司郑州市分公司,河南郑州450000

出　　处：《数字技术与应用》2021年第5期79-81,共3页Digital Technology & Application

基　　金：2020年河南省档案局项目(2020-X-12),项目名称:人工智能在电子图像自动分类中的应用研究。

摘　　要：人工智能技术正在对人类生活产生着深刻的影响,随着数字档案信息资源的种类和数量快速增长,传统的档案信息资源检索已无法满足目前的社会需求。本文使用Faster R-CNN深度学习网络模型对数字图像进行目标检测,并将数字图像及相关参数存储于非结构化数据库中,使检索能够搜索出和用户录入的搜索内容最为相似的结果,从而大大减少用户检索的时间,提高检索效率。Artificial intelligence technology is having a profound impact on human life.With the rapid growth of the types and quantity of digital archival information resources,the traditional archival information resources have been unable to meet the current social needs.In this paper,the deep learning network model named Faster R-CNN is used for target detection,and the digital images and related parameters are stored in an unstructured database,so that the search results can find the most similar results with the search content entered by the user.Thus,the retrieval time of users is greatly reduced and the retrieval efficiency is improved.

关 键 词：深度学习 数字图像档案 智能检索 卷积神经网络 Faster R-CNN 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]