| 基于服务机器人的人体追踪技术的研究 |
| |
| 引用本文： | 汤辰,万衡. 基于服务机器人的人体追踪技术的研究[J]. 电子器件, 2015, 38(6) |
| |
| 作者姓名： | 汤辰 万衡 |
| |
| 作者单位： | 上海应用技术学院 |
| |
| 摘 要： | 在计算机视觉领域中，对于人体的识别追踪等技术的广泛应用而受到了大量学者的关注与研究。现有通过检测人体骨架，提取其深度信息，根据骨骼点坐标来进行追踪。然而实验发现无论是通过微软SDK或者Openni来提取人体骨骼特征都会出现丢失和误判断问题，以及在多人干扰下就会出现识别混乱等情况。针对以上问题，提出了结合激光扫腿的纠正方法，有效地解决了传统方法中不能在跟踪失败时自恢复问题，并且在服务型机器人的平台中实验得到验证。
|
| 关 键 词： | 机器视觉 人体追踪 骨架 激光扫腿 |
The human body tracking technology research based service robot |
| |
| Abstract: | In the field of computer vision,As to the Recognition Technology and Tracking Technology are widely used for human body,a large number of scholars research these.By detecting the human body skeleton,extract the depth information,according to the skeleton point coordinates to track.Although the experiment find a problem that the skeleton feature may lost and false judgment by Microsoft SDK or Openni,and appear to identify chaos under many interference.To solve above problems,the correction method of laser scanning legs is proposed,effectively solve the problem that it can not restore when tracking failure in the traditional method,And the experiment in the service robot platform has been verified. |
| |
| Keywords: | Machine vision KeywordsThe human body tracking Skeleton Laser scanning legs |
|
| 点击此处可从《电子器件》浏览原始摘要信息 |
|
点击此处可从《电子器件》下载免费的PDF全文 |
|
