文献类型：期刊文献

中文题名：综合分块颜色体积直方图和LGWBPs的图像检索

英文题名：Iimage Retrieval of Integrated Block Color Volume Histogram and LGWBPs

作者：刘芳辉[1];郭慧[1]

机构：[1]华东理工大学机械与动力工程学院,上海200237

年份：2017

卷号：38

期号：3

起止页码：388

中文期刊名：图学学报

外文期刊名：Journal of Graphics

收录：北大核心:【北大核心2014】；CSCD:【CSCD_E2017_2018】；

语种：中文

中文关键词：多特征;图像检索;颜色体积直方图;局部Gabor二元模式

外文关键词：multiple feature; image retrieval; block color volume histogram; local Gabor wavelets binary patterns

摘要：针对多特征融合提高图像检索效果问题,提出一种综合分块颜色体积直方图和局部Gabor二元模式(LGWBPs)的图像检索方法,有效地提取了图像的颜色和纹理特征,为用户提供了精准的图像检索结果。首先,将4×4分块图像进行重叠九分块,并提取每个分块的颜色体积直方图特征;然后用定义的LGWBPs算子统计每个子块信息并提取图像的LGWBPs直方图特征;最后综合提取的两个特征进行相似性度量。实验结果表明,该方法有效地提高了检索的精准率,改善了检索结果的排序值并具有较好的旋转不变性、抗噪性和不变特征的鲁棒性。
A novel image retrieval method based on integrated block color volume histogram and local Gabor wavelets binary patterns(LGWBPs)is proposed to solve the problem of multiple feature fusion to improve the image retrieval efficiency,The proposed method can effectively extract the color and texture features of the image,and provide users with accurate image retrieval results.Firstly,the4×4block image is divided into nine blocks and the color volume histogram features of each block are extracted.Then,LGWBPs histogram features are calculated by using the defined LGWBPs operator,and the extracted LGWBPs histogram features are extracted.Finally,the two features are integrated to measure the similarity.The experimental results show that the proposed method can improve the accuracy of retrieval,improve the ranking of retrieval results and has good rotation-invariance,anti-noise and robustness of invariant features.