Improving CBIR accuracy using convolutional neural network for feature extraction
Improving CBIR accuracy using convolutional neural network for feature extraction
复制标题
使用卷积神经网络进行特征提取提高 CBIR 精度
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
M. A. Shah
中科院分区：
文献类型：
--
作者：
Amjad Shah;Rashid Naseem;Sadia;Shahid Iqbal;M. A. Shah
由于多媒体含量及其视觉复杂性的快速增长，基于内容的图像检索（CBIR）成为非常具有挑战性的TAK。从图像查询到相关图像的检索，CBIR具有不同的阶段。但是，图像的特征提取是重要阶段之一。最近，由于图像中提取特征的能力，卷积神经网络（CNN）在计算机视野领域显示出良好的结果。本文在CBIR系统中介绍了从图像中提取特征的CNN。欧几里得距离用于使用提取的特征之间的查询和存储图像之间的关联。使用精度评估拟议工作的性能。与现有作品相比，提议的工作显示出改善的结果。
Content Based Image Retrieval (CBIR) becomes a very challenging taks due to the rapid growth in multimedia content and its visual complexity. From query by image to retrieval of relevant images, CBIR has different phases. However, features extraction of images is one of the important phases. Recently Convolutional Neural Network (CNN) shows good results in the field of computer vision due to the ability of extraction features from the images. This paper introduces CNN for features extraction from images, in CBIR system. Euclidean distance is used for association among query and stored images using the extracted features. Performance of the proposed work is evaluated using precision. The proposed work shows improved results as compared to the existing works.