提出了一种运用行为规则库和综合使用听觉信息、视觉信息来完成目标搜索的策略。该策略首先由机器人的传声器阵列和双目立体视觉摄像机采集环境信息,然后通过行为控制中心决定执行规则库中的动作完成目标搜索,从而避免使用机器人的单一感官信息造成的难题。行为规则库由4条和机器人听觉模块和视觉模块相关的规则组成。听觉模块基于五元传声器阵列实现目标空间定位,视觉模块基于颜色和特征组合实现目标识别。系统在室内环境下进行了测试,实验结果证明机器人可以判定并且搜索到目标,结果证明该方法具有实时实现的有效性和应用性。
A strategy that uses a behavior rule base and comprehensively utilizes hearing information and vision information to search a target is proposed in this paper. The strategy is described as below. Firstly, the environment information acquisition is realized using the robot's 5-microphone array and binocular camera, and then the behaviors in the behavior rule base are executed according to a behavior control center to achieve the target search. The behavior rule base is composed of four rules relevant to the robot hearing module and the vision module. The hearing module target localization is realized based on the 5-microphone array, and the vision module target recognition is realized based on the combination of color and characteristics. The system was tested in a room environment, and the experimental results show that the robot is capable of judging and searching the target. The method is efficient and practical enough for real-time operation.