论文部分内容阅读
提出并实现了一种用于移动机器人实时自主导航系统的结构光视觉传感器。详细介绍了该视觉传感器的工作原理、数学模型、标定方法以及视觉图像的处理方法。这种主动式方法避开了立体匹配过程,实现了障碍物位置信息的快速探测。利用传感器几何约束,有效地剔除了的视觉图像中干扰的信息。该视觉传感器已成功应用于室内移动机器人避障与导航,说明该视觉方法是有效、可靠的。
A structured light vision sensor for real-time autonomous navigation system of mobile robot is proposed and implemented. The working principle, mathematical model, calibration method and visual image processing method of the visual sensor are introduced in detail. This proactive approach avoids the stereo matching process and enables fast detection of obstacle location information. Using sensor geometry constraints, the information of disturbances in the visual image is effectively rejected. The visual sensor has been successfully applied to indoor mobile robot obstacle avoidance and navigation, indicating that the visual method is effective and reliable.