CAJ | 학술논문

介绍了一种基于肤色与手形的动态人手跟踪方法。根据检测过程中所用到的人手基本特征，提出了一种基于色度一饱和度自适应阂值肤色分割与成对几何直方图（PGH）匹配的人手图像检测方法。为了克服跟踪过程中背景色干扰的问题，实现了基于Kalman滤波器的预测跟踪。实验结果表明，该动态人手跟踪方法计算简便、实时性好、跟踪精度高，可应用在大多数复杂场景。
개소료일충기우부색여수형적동태인수근종방법。근거검측과정중소용도적인수기본특정，제출료일충기우색도일포화도자괄응애치부색분할여성대궤하직방도（PGH）필배적인수도상검측방법。위료극복근종과정중배경색간우적문제，실현료기우Kalman려파기적예측근종。실험결과표명，해동태인수근종방법계산간편、실시성호、근종정도고，가응용재대다수복잡장경。
This paper presents an approach of moving hand tracking based on skin-color and hand shape. It proposes a method to detect moving hand based on chromatic adaptive thresholding and PGH matching. In order to solve the problem of the disturbances from background color, it achieves a predictive tracking method based on Kalman filter. The results show that the method is simple and accurate, and can be used in most complex scenes.