基于多传感器的篮球机器人目标识别与定位研究  

作　　者：鲁晨奇 陈万米[1] 胡晨潇 

机构地区：[1]上海大学机电工程与自动化学院,上海200444

出　　处：《工业控制计算机》2019年第1期64-66,69,共4页Industrial Control Computer

摘　　要：介绍了篮球机器人比赛规则及篮球机器人系统结构层次,明确感知子系统的主要任务要求是对目标进行识别与定位。接着,针对目标识别过程易受光照条件影响的问题,决定采用了HSV颜色空间。根据球的颜色特征和轮廓形状特征,提出了基于颜色和轮廓外形的目标识别算法。在目标定位方法中,提出了"长距离使用Kinect传感器,短距离使用激光传感器"的定位策略。最后,综合上述研究成果,在篮球机器人通过目标识别实验进行检验,成功实现了目标的快速、准确识别。This paper introduces the basketball robot game rules and the structure level of the basketball robot system.It is clear that the main task requirement of the perception subsystem is to identify and locate the target.Next,the HSV color space was decided for the problem that the target recognition process was susceptible to illumination conditions.According to the color characteristics and contour shape characteristics of the ball,a target recognition algorithm based on color and contour shape is proposed.In the target positioning method,a positioning strategy of using a long-distance using a Kinect sensor and using a laser sensor for a short distance is proposed.Finally,based on the above research results,the basketball robot successfully tested the target recognition experiment and successfully achieved the target's rapid and accurate recognition.

关 键 词：篮球机器人 目标识别 目标定位 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]