【摘要】 为了适应机器人足球视觉系统图像采集设备从单个到多个的改变,提高系统的跟踪成功率和速度,以信息融合技术为基本思路,提出了一种基于多传感器多目标跟踪的半自主足球机器人视觉跟踪方法.各跟踪模块对覆盖区域的目标进行航迹关联、滤波后再对各传感器信息进行融合得到所有目标单一的航迹.实验表明,该方法具有简单、有效,满足系统实时性要求等特点.更多还原

【Abstract】 A tracking approach based on multisensor-multitarget tracking of the basic theory of information fusion is introduced to fit the change from singlesensor to multisensor of FIRA MiroSot game’s vision system, and to improve the reliability and the processing speed of the system. After track correlation and filtering for every senor data, the tracks of target for every sensor are obtained. An information fusion method for all these target tracks is presented. Experimental results show that the method is succinct and effective.更多还原