Wide Angle Tracking System for Mobile Robots

最近研究了一种新的移动机器人导航和控制技术。该技术使用广角镜头对指定目标进行成像，并通过图像检测和识别来确定机器人相对于目标的位置，从而为导航控制提供依据。这种传感器也可以单独使用或与其他传感器一起使用，为静止或移动的物体提供避障功能。在层次控制结构中，研究了更高层次的路径跟踪和区域填充问题。对于非结构化和不断变化的环境，导航控制问题是至关重要的。本研究针对视觉控制问题展开。特别提出了一种高速视觉跟踪方法。研究了一种使用多数据和处理路径的二值图像处理的硬件结构。通过这种方式，我们希望实现视觉跟踪速率，这将允许基于视觉的控制器用于各种移动机器人。

A new technique for mobile robot navigation and control has recently been investigated. The technique uses a wide angle lens to image specified targets and uses image detection and recognition to determine the position of the robot with respect to the targets thus providing a basis for navigation control. This sensor may also be used alone or with other sensors to provide obstacle avoidance for stationary or moving objects. The higher level problems of path following and region filling are also included in this research in a hierarchical control structure. For unstructured and changing environments, the navigation control problem is crucial. This research is directed toward the vision control problem. In particular, a high speed vision tracking method is proposed for this research. A hardware architecture using binary image processing with multiple data and processing paths will be investigated. In this manner, we hope to achieve vision tracking rates which will permit a vision based controller to be used for a wide variety of mobile robots.