论文部分内容阅读
本文利用彩色图像对移动机器人进行定位，将颜色由 Ｒ Ｇ Ｂ 空间转换到 Ｈ Ｓ Ｉ空间中，颜色成为图像的唯一特征，简化了图像的特征抽取，大大提高了图像的处理速度和特征抽取的鲁棒性．在结构上采用全局静态视觉，避免了对摄像机的控制，并使得定位时在环境中只需一个着色标记．在移动机器人运动过程中，利用动态窗口进一步减少数据处理量．考虑视觉系统的延时，推广卡尔曼滤波完成视觉与推算航行数据的融合．该定位方法已成功地应用于我们研制的移动机器人控制系统中，效果良好．
In this paper, a color image is used to locate a moving robot. The color is converted from R G B space to H S I space, and the color becomes the unique feature of the image, which simplifies the image feature extraction and greatly improves the image processing speed and feature extraction Robustness. The use of global static vision on the structure eliminates the need for camera control and enables only a single color mark in the environment when positioning. In the mobile robot movement process, the use of dynamic windows to further reduce the amount of data processing. Considering the delay of vision system, the extended Kalman filter is used to complete the fusion of vision and projected navigation data. The positioning method has been successfully applied to our development of mobile robot control system, the effect is good.