Double-robot dynamic stereo vision algorithm based on UKF and its application in moving target tracking

作者：Dong Huiying^*; Lu Xue; He Yuqing; Chen Shengfu; Han Jianda

来源：Advances in Information Sciences and Service Sciences, 2012, 4(18): 263-270.

DOI：10.4156/AISS.vol4.issue18.32

• 3D moving targets
• Formation control
• Ground robot
• Moving target tracking
• Moving targets
• Research directions
• Stereo vision algorithms
• Stereo-vision system
• UKF
• Vision sensors

Cooperation between air and ground robots has gradually been one of the latest research directions in robotics, wherein position of ground robot assisted by air robots is a key problem to be solved. This paper, therefore, mainly focuses on the problem of high precise 3D moving target observation. Firstly, a new so-called dynamical stereo-vision system, composed of multiple air robots with vision sensor mounted, is proposed and the corresponding observation algorithm is studied. Secondly, in order to improve the observation precision and realize persistent observation, UKF algorithm is utilized to form a new method for moving target observation and tracking. Subsequently, formation control algorithm is designed to make the air robots keep tracking the interested moving target. Finally, experiments are conducted with respect to a new-designed platform, called indoor multiple-rotor-craft-robot system, to verify the feasibility and validity of the study.

• 出版日期2012