摘要： 存储于计算机上日益庞大的数字图像存在语义表达隐含、存储孤立、检索不便等缺点，导致图像应用效率低。该文提出基于本体的图像语义标注与检索模型ImageQ，其具有层次及模块化特点且与领域本体相独立，支持领域本体进化和系列推理深度不同的图像检索方法。基于ImageQ的实验系统运行结果表明，该模型的图像管理性能良好，实现了图像智能检索功能。

关键词: 领域本体, 图像标注, 图像管理, 智能检索

Abstract: The increasingly huge volume of digital images stored in computer are characterized by implicit semantic, disconnected storage and inconvenient in retrieval, thus result in low efficiency in image using. This paper proposes the ImageQ which is an ontology-based annotation and retrieval model for image semantic. It features the layered and modular architecture and independent of the domain ontology, supports domain ontology evolution and several image retrieval methods in different depth of reasoning. The running results of a experimental system based on ImageQ show that the model well performed in image management and intelligent retrieval.

Key words: domain ontology, image annotation, image management, intelligent retrieval

中图分类号: 

• TN919.81