提出一种基于本体的图像情感语义识别系统,使情感识别过程易于理解。针对人类在图像情感语义理解过程中存在的个性化和群体化现象,系统使用多层次的情感推理模型。在阐述系统的设计思想和各模块的工作机理的同时,着重介绍了通过语义网规则语言表示的多层次情感推理模型的组织结构和表示方法以及推理机的工作原理。仿真实验结果表明了在图像情感语义识别及推理过程中引入本体概念的可行性。
An image affective semantic recognition based on ontology and its reasoning model is introduced,the model can make the reasoning procedure more intelligible.By researching population and individuation phenomena in image comprehension,this model is a multi-layer reasoning model.The designing thoughts and working mechanism of the system is introduced,the structure of the multilayer reasoning model represented by SWRL(semantic web rule language) as well as the principle of the reasoning machine are discassed in detail.The experimental result shows that the approach using ontology in image affective semantic recognition is viable.