基于深度卷积神经网络的图像检索算法研究  

Image retrieval based on deep convolutional neural network

作　　者：刘海龙 李宝安[2] 吕学强[1] 黄跃[3] 

机构地区：[1]北京信息科技大学网络文化与数字传播北京市重点实验室,北京100101 [2]北京信息科技大学计算机学院,北京100101 [3]首都医科大学宣武医院,北京100053

出　　处：《计算机应用研究》2017年第12期3816-3819,共4页Application Research of Computers

基　　金：网络文化与数字传播北京市重点实验室开放课题(ICDD201504);国家自然科学基金资助项目(61271304);北京成像技术高精尖创新中心项目(BAICIT-2016003);2014年度国家社会科学基金委托课题(14@ZH036)

摘　　要：为解决卷积神经网络在提取图像特征时所造成的特征信息损失,提高图像检索的准确率,提出了一种基于改进卷积神经网络LeNet-L的图像检索算法。首先,改进LeNet-5卷积神经网络结构,增加网络结构深度;然后,对深度卷积神经网络模型LeNet-L进行预训练,得到训练好的网络模型,进而提取出图像高层语义特征;最后,通过距离函数比较待检图像与图像库的相似度,得出相似图像。在Corel数据集上,与原模型以及传统基于SVM主动学习图像检索方法相比,该图像检索方法有较高的准确性。经实验结果表明,改进后的卷积神经网络具有更好的检索效果。To solve the problem that the loss of image feature information and improve the accuracy of image retrieval,when the convolutional neural network( CNN) was used to extract the feature information of the image,this paper proposed an image retrieval algorithm based on improved convolutional neural network Le Net-L. First,it improved Le Net-5 convolution neural network structure,increased depth of network structure. Then,it pre-trained the deep convolutional neural network Le Net-L to extract the high-level semantic features. At last,it obtained the similar images by distance function between the image being retrieved and the one in image database. In Corel dataset,compared with the original model method and the traditional image retrieval method based on SVM and active learning,the proposed method had a higher accuracy. The experimental results show that the improved convolutional neural network has a better retrieval effect.

关 键 词：图像检索 卷积神经网络 特征提取 深度学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]