陈伟, 吴涛, 李政, 贺汉根. 基于粒子滤波的单目视觉SLAM算法[J]. 机器人, 2008, 30(3): 242-247,253.
引用本文: 陈伟, 吴涛, 李政, 贺汉根. 基于粒子滤波的单目视觉SLAM算法[J]. 机器人, 2008, 30(3): 242-247,253.
CHEN Wei, WU Tao, LI Zheng, HE Han-gen. A Monocular Vision SLAM Algorithm Based on Particle Filter[J]. ROBOT, 2008, 30(3): 242-247,253.
Citation: CHEN Wei, WU Tao, LI Zheng, HE Han-gen. A Monocular Vision SLAM Algorithm Based on Particle Filter[J]. ROBOT, 2008, 30(3): 242-247,253.

基于粒子滤波的单目视觉SLAM算法

A Monocular Vision SLAM Algorithm Based on Particle Filter

  • 摘要: 针对携带有单目摄像机和码盘的微小机器人的定位与建图问题,提出了基于粒子滤波的SLAM(同时定位与建图)算法.从摄像机中提取图像特征点,并在图像序列中加以匹配,根据相应时刻的摄像机位姿计算得到对应的环境标志点坐标;机器人的大致位姿估计由码盘运动模型获得.在机器人移动过程中,环境标志点的观测信息和码盘信息通过粒子滤波相融合,从而提高了机器人定位的精度,同时也得到了更为准确的环境标志点坐标.仿真实验结果表明本算法有效、可靠.

     

    Abstract: To deal with the localization and mapping problem of miniature robot equipped with monocular camera and encoders,an SLAM(simultaneous localization and mapping) approach based on particle filtering is presented.The image features are detected from camera and matched among successive frames,and then coordinates of the corresponding features in environments are computed according to the camera poses.The rough estimates of robot poses are obtained from the encoder motion model.During robot motion,the information from feature observations is fused with that from the encoders by particle filter,so location precision of the robot is improved and also more exact coordinates of the features are gotten.The reliability and efficiency of the method are proved by simulation experiments.

     

/

返回文章
返回