基于图像的移动机器人视觉伺服

Image-based visual servoing for mobile robots

摘要:

为了更简便地将基于图像的视觉伺服运用于移动机器人，避免采用近似线性输入输出反馈控制模型中近似与假设的情况，提出了3种使用极线几何与三角几何相结合的方法，此类方法不需要预先知道三维场景的结构知识。实验仿真结果证明了该方法的有效性。

Abstract:

To achieve Image-Based Visual Servoing(IBVS) for mobile robots more easily，and avoid using an approximate input-output linear feedback which has many assumptions，three methods combining epipolar geometry with triangle geometry were proposed. The proposed approaches did not need any knowledge of 3-D scene geometry. Simulation results validate the effectiveness of the proposed approach.