摘要
提出一种基于本体的图像检索方法。该方法结合特定领域专家知识和对象例图,采用视觉对象本体来描述图像内特定对象的视觉特征,从而构建该领域包含视觉描述的知识库。在检索过程中,利用知识库内的对象的视觉本体描述和目标图像库内的图像低层特征相匹配执行图像检索任务,从而实现在高层次语义上的图像检索。实验结果表明了该方法的有效性和可行性,并在一定程度上缩小了视觉低层特征同图像高层语义的"鸿沟"。
An image retrieval method using ontology is presented in the paper.In the method, combining the knowledge of experts in special field and the object case image, using visual object ontology to describe visual features of special target in the image, we constructed the knowledge base which contains visual description of objects.In order to achieve image retrieval at a high-level semantic in the retrieval process, we implement the retrieval mission by matching description of object which was described by visual object ontology in knowledge base and low-level features of target images.The results of the experiment not only show the efficiency and feasibility of the method, but also have significance for reducing the＂gap＂between high-level image semantics and low-level vision features of the pictures.
出处
《微计算机信息》
2010年第8期210-211,214,共3页
Control & Automation
关键词
图像检索
本体
图像语义
语义对象
Image Retrieval
Ontology
Image Semantic
Semantic Object
