Abstract: A new image retrieval method using high-level semantics is presented in this paper.First，the proposed approach employs a segmentation algorithm based on color and space to divide images into regions，and low-level feature for the color，shape，position，and texture of each region are subsequently extracted.The cluster image object using these features and then the semantic feature is gained；FCM is used to cluster image.In image retrieval，the querying image is compared to cluster center，then retrieved in the class with the minimal distance.The experiment results show that the proposed approach has an excellent precision and has reduced the “semantic gap” between the visual feature and semantic visual.

Key words: image retrieval, image segmentation, FCM, Clustering

摘要： 提出了一种基于高层语义的图像检索方法，该方法首先将图像分割成区域，提取每个区域的颜色、形状、位置特征，然后使用这些特征对图像对象进行聚类，得到每幅图像的语义特征向量；采用模糊C均值算法对图像进行聚类，在图像检索时，查询图像和聚类中心比较，然后在距离最小的类中进行检索。实验表明，提出的方法可以明显提高检索效率，缩小低层特征和高层语义之间的“语义鸿沟”。

关键词: 图像检索, 图像分割, 模糊C均值, 聚类