摘要
基于内容的图像检索是当前多媒体信息检索的热点之一。基于内容的图像检索技术是根据对图像内容(特征)的描述和提取,在图像库中找到具有指定内容(特征)的图像。本文对图像颜色特征和纹理特征的提取、相似性度量等基于内容的图像检索的关键技术进行了分析和研究,并在此基础上,提出了一个基于颜色特征和纹理特征的图像检索算法并验证了其有效性。该算法采用HSV颜色空间的直方图作为颜色特征向量,采用灰度共生矩阵的四个纹理特征:能量、熵、惯性矩和相关性构成纹理特征向量,采用欧氏距离进行相似性度量。实验结果表明,该算法实现的系统具有良好的图像检索功能。
Content-Based Image Retrieval （CBIR） is one of the most active hot spots in the current research field of multimedia retrieval. According to the description and extraction of visual content （feature） of the image, CBIR aims to find images that contain specified content （feature） in the image database. In this paper, several key technologies of CBIR such as the extraction of the color and texture features of the image, as well as the similarity measures are investigated. On the basis of the theory research, an image retrieval system based on color and texture features is designed, which uses Histogram based on H$V color space as color feature vector, uses four of the features of Co-occurrence Matrix to construct texture vector which are Energy, Entropy, InertiaQuadrature and Correlation, and uses Euclidean distance for similarity measure. Experiments results show that this CBIR system is efficient in image retrieval.
关键词
图像检索
基于内容
特征提取
相似性度量
image Retrieval
content-based
feature extraction
similarity measures
