基于语义分类的文物图像标注研究

由于图像数据中普遍存在的“语义鸿沟”问题,传统的基于内容的图像检索技术对于数字图书馆中的图像检索往往力不从心。而图像标注能有效地弥补语义的缺失。文中分析了图像语义标注的现状以及存在的问题,提出了基于语义分类的文物语义标注方法。算法首先通过构建一个Bayes语义分类器对待标注图像进行语义分类,进而通过在语义类内部建立基于统计的标注模型,实现了图像的语义标注。在针对文物图像进行标注的实验中,该方法获得了较好的标注准确率和效率。

Research on Semantic Annotation to Cultural Relic Images Based on Semantic Classification

Because of the "semantic gap" which is often encountered in the image data,traditional CBIR technology can't deal with the problem of image retrieval in digital libraries sometimes.Image annotation can counterbalance semantic loss.In this paper the research situation of the problems existed in image semantic annotation is overviewed.An image annotation method based on the semantic classification is put forward.The algorithm classify images which are to be annotated firstly by the Bayes semantic classifier constructed,then the semantic annotation to the images are got by establishing a statistical annotation model within those semantic classes.Experimental results on a large collection of cultural relic images have shown the effectiveness and robustness of proposed algorithms.