首页 | 本学科首页   官方微博 | 高级检索  
     

卷积特征图融合与显著性检测的图像检索
引用本文:聂一亮,杜吉祥,杨麟. 卷积特征图融合与显著性检测的图像检索[J]. 华侨大学学报(自然科学版), 2018, 0(6): 937-941. DOI: 10.11830/ISSN.1000-5013.201706028
作者姓名:聂一亮  杜吉祥  杨麟
作者单位:华侨大学 计算机科学与技术学院, 福建 厦门 361021
摘    要:针对基于深度学习的图像检索提取特征往往包含了复杂的背景噪声,导致图像检索的精确率并不高的问题,提出一种特征图融合与显著性检测的方法.首先,训练用于分类的深度卷积神经网络模型.然后,并将图像卷积之后的特征图谱进行融合,得到图像的显著性区域.最后,通过计算图像显著性特征的余弦距离来进行检索.实验结果证明:相比目前主流的方法,文中方法能够有效提高检测精度,且鲁棒性较高.

关 键 词:图像检索  特征图融合  显著性检测  卷积神经网络

Image Retrieval Based on Convolution Feature Map Fusion and Saliency Detection
NIE Yiliang,DU Jixiang,YANG Lin. Image Retrieval Based on Convolution Feature Map Fusion and Saliency Detection[J]. Journal of Huaqiao University(Natural Science), 2018, 0(6): 937-941. DOI: 10.11830/ISSN.1000-5013.201706028
Authors:NIE Yiliang  DU Jixiang  YANG Lin
Affiliation:College of Computer Science and Technology, Huaqiao University, Xiamen 361021, China
Abstract:Based on an in-depth learning of image retrieval, the features extracted usually contained the complicated background noises, which resulted in a low level of accuracy in image retrieval. The methods of feature map fusion and saliency detection are proposed in this paper. The method firstly trained deep convolutional neural network model used in image classification, and then fused the features of maps after image convolution in order to obtain the salient region of retrieved images. Finally, the retrieved images are calculated using the cosine distance of the salient features. The experiment shows that the proposed methods are able to effectively improve the accuracy of retrieval and that the robustness is relatively high, compared to the current mainstream methods.
Keywords:image retrieval  feature map fusion  saliency detection  convolutional neural network
本文献已被 CNKI 等数据库收录!
点击此处可从《华侨大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《华侨大学学报(自然科学版)》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号