Intelligent Perception Based on Knowledge Base and Its Application

Aiming at the problem that existing virtual assembly interfaces lack scenario analysis capabilities and can´t effectively perceive user´s intentions, the scene knowledge model and the user knowledge model are proposed in this paper. We establish scene knowledge base and user knowledge database according to the scene knowledge model and the user knowledge model respectively. To solve the ambiguity problem of the intention of user interaction, expert system of uncertain reasoning is used for understanding the specific semantics of user´s actions. The innovations of this work include: (1) we construct the scene knowledge model and the user knowledge model separately. We combine scene knowledge base and user knowledge database closely to build rule base, (2) because of the fuzziness of the user´s action, the method of uncertainty reasoning based on credibility is adopted. The experimental results show that the proposed method can perceive user´s intentions accurately.