视觉伺服移动机器人中的图像处理研究

Study of Image Processing of Visual Servoing Mobile Robot

从机器人彩色视觉的形成机理入手 ,分析了影响视觉的主要因素 ,其中包括光照条件与成像元件等 ,设计了一种基于知识融合的视觉图像处理方法 ,将其应用于机器人的彩色视觉后 ,增加了机器人视觉系统的鲁棒性 .研究表明 ,该方法可以在环境光场有变化或光场分布不均匀的情况下快速、准确地计算出移动机器人的位置与姿态 ,定位误差不超过 4× 4的像素点范围 ,角度误差在±2° ,因此满足了视觉伺服系统性能指标的要求 .

Starting with the mechanism of mobile robot's color vision, main influence factors of vision including the condition of illuminance and the imaging element are analyzed. The method of vision image processing is designed based on knowledge fusion. The vision system of mobile robot has more robust by using the image processing. The method can exactly and rapidly work out the pose and the position of mobile robot in variational or asymmetrical illumination conditions. The location error is less the 4×4 pixels and the angle error is in the range of ±2 degrees.