According to the robot based on their own collection of complete visual image barriers identification process, because barriers image features of the complex, the obstacles visual characteristics identification with ambiguity, will occur recognition error. Put forward based on the semantic ambiguity constraint natural model robot vision image barriers recognition system design method, has established a set of complete hardware module design scheme, and natural language constraints design method of software realize the visual obstacle robot ambiguity ambiguity eliminate. The experimental results show that the system in intelligent robot obstacle recognition good eliminate the existing ambiguity, improve the recognition accuracy%针对机器人在完成基于自身采集的视觉图像障碍识别过程中,由于障碍图像特征复杂,障碍视觉特征带有识别二义性,会发生识别错误；提出了一种基于歧义约束自然语义模型的机器人视觉图像障碍识别系统设计方法,建立了一套完整的硬件模块设计方案,并用约束自然语言软件设计方法实现了机器人视觉障碍二义性歧义消除;实验结果证明,该系统在机器人视觉智能障碍识别方面很好地消除了存在的二义性,提高了识别的准确度.
展开▼
