计算机与数字工程Issue(9)：2276-2280,2285,6.DOI:10.3969/j.issn.1672-9722.2019.09.033
基于深度学习的图像检索研究∗
Research on Image Retrieval Based on Deep Learning
摘要
Abstract
Firstly,deep learning is a new field in machine learning research. It is widely applied to image recognition and has achieved good results. Secondly,how to retrieve target image from large-scale image data is becoming the focus and difficulty of mul?timedia research. So in this paper,based on the depth study of image retrieval technology,the depth of the convolutional neural net?work model for feature extraction is used,the characteristics are binarized,and the network loss function is modified,and then the binary operator is applied to image retrieval system,the efficiency of image retrieval is improved. The key step of image retrieval is to make accurate image feature extraction and efficient feature matching.关键词
深度学习/深度卷积神经网络/图像检索/二进制描述符Key words
deep learning/deep convolution neural network/image retrieval/binary descriptor分类
信息技术与安全科学引用本文复制引用
杨馥溢,何嘉..基于深度学习的图像检索研究∗[J].计算机与数字工程,2019,(9):2276-2280,2285,6.基金项目
四川省科技计划项目(编号:2012GZ0111)资助. （编号:2012GZ0111）
