Abstract: This paper presents and realizes a new image retrieval method by combining the information of color connected area and its features of texture. The image is distributed into several parts, the color connected areas in the image is computed, the primitive co-occurrence matrix of four colors’ corresponded with the connected area of each color is extracted, the image retrieval on the basis of the content is realized through the use of the feature similar function which is designed according to these features. The result of the experiment shows that this method effectively combines the information of image’s textures and its color constitution as well as its distribution which has a good effect.

Key words: image retrieval, co-occurrence matrix, image texture, texture extraction

摘要： 提出一种结合图像颜色连通区域信息及其纹理特征的图像检索新方法。对图像进行分块，确定图像中的颜色连通区域集，提取图像中各颜色连通区域对应的4种颜色基元共生矩阵特征，针对该特征设计的图像相似性度量函数实现基于内容的图像检索。实验结果表明，该方法能有效地结合图像的纹理信息及其颜色构成和分布信息，具有良好的检索效果和性能。

关键词: 图像检索, 共生矩阵, 图像纹理, 特征提取

CLC Number: 

• TP391.41