人眼检测在机械臂中的应用

Application of Human Eye Detection in Robot Arm

针对没有手臂的残疾人士,设计一款基于人眼控制的机械臂。该方法主要采用AdaBoost级联分类器检测人脸,通过加入感兴趣区域的特征点检测进行改善降低误检测,然后在人脸范围内检测人眼输出矩形波,机械臂通过处理获取的矩形波完成一些列操作。实验结果表明该方式可以很好地实现通过检测眼睛输出矩形波控制机械臂,并且具有好的实时性和鲁棒性。

A mechanical arm based on human eye control for disabled persons without arms is proposed.The method adopts mainly AdaBoost algorithm to detect human faces.And the feature point detection in ROIregion is introduced to reduce false detection ratio.When human eye in the face range is detected,rectangle wave is exported.The mechanical arm completes a series of operations according to the rectangle wave acquired.The experimental results show the method proposed in this article is feasible and has strong robustness.