Abstract: This paper proposes a color image retrieval algorithm using main color, clustering index. It quantifies the image using MPEG-7 visual content description, selects the main color and its percentage as the color characteristics, establishes the clustering index database based on the main colors. The similarity between image is computed by using the color characteristics and image fast retrieval is obtainable by using the clustering index. Experimental results show that the proposed image retrieval is more efficient in retrieving the user interested images with faster retrieval speed.

Key words: image retrieval, color, HSV, feature extraction, cluster index

摘要： 提出了一种基于主颜色、聚类索引表的彩色图像检索算法。应用MPEG-7视觉内容描述对彩色图像进行量化处理，选取图像的主颜色及其所占百分率作为颜色特征，根据主颜色组合建立聚类索引数据库。利用上述主颜色特征计算图像间的相似度，利用聚类索引表对图像进行聚类和快速检索。实验表明，该算法能够准确和高效地检索出用户所需的彩色图像，具有较快的检索速度。

关键词: 图像检索, 颜色, HSV, 特征提取, 聚类索引