【摘要】 基于内容的图像检索技术是对图像的物理内容为加工对象的检索技术之一,主要实现方式包括基于颜色、纹理、形状、空间位置和语义等。其中基于颜色的图像检索发展最为成熟,而基于语义的检索则尚处于探讨、研究阶段。基于内容检索和基于文本检索在数字图书馆中可以融合共同提供检索服务。Google为这一尝试提供了在后控阶段的有效案例,而真正的实现两者的融合是在预处理阶段。两者结合在数字图书馆中的应用是可行的,相信能够提供更好的图像检索服务。更多还原

【Abstract】 Content-based image retrieval is information retrieval technique that takes the physical content of images as its objects. It mainly includes retrieval with color, texture, shape, spatial and semantics, etc. Among the above methods, color-based method is the most mature while semantic image retrieval is under pilot studies at present. Both content-based and text-based images can be incorporated into digital libraries to provide information retrieval services. Google has provided a successful case for the effort in post-control stages. The real combination, however, is at the pre-control stage. It is feasible to combine the both in digital libraries and believed to provide better image retrieval services.更多还原