论文部分内容阅读
研究了基于内容的图像检索系统中的目标描述模型的建立方法。首先指出了目标描述模型是图像检索的关键技术 ,然后分析了MPEG 7草案中有关多媒体描述的基本术语、描述机制和MPEG 7的应用框架 ,最后基于MPEG 7提出了一种适合于图像检索的目标描述模型。该模型对提取出的多种视觉特征和相应的表示方法采用了分层结构。模型满足用户对所需特征进行不同级别检索的要求。
The method of establishing the target description model in the content-based image retrieval system is studied. First of all, it points out that the target description model is the key technology of image retrieval, and then analyzes the basic terminology, description mechanism and the application frame of MPEG 7 in MPEG 7 draft, and finally proposes a suitable image retrieval target based on MPEG 7 Describe the model. The model employs a hierarchical structure for the various extracted visual features and corresponding representations. The model satisfies the user’s requirements for different levels of retrieval of desired features.