动态手势轨迹跟踪与HMM模式识别

Dynamic Hand Gesture Trajectory Tracking and Hidden Markov Model Recognition

为了理解用户手势的含义,该文首先利用摄像头捕获连续运动的手势,随后将其进行部分图像的预处理操作,利用手势分割技术将手势部分分割出来,并提取手势的质心以及面积的大小作为特征向量,用于建立隐马尔可夫模型的参数,最终形成一个能理解用户手势操作的系统,并对其进行应用。

To understand the meaning of users' hand gestures, first of all, this article uses cameras to capture a series pictures of moving hands and then predo these pictures by using the technology of hand gestures segment to identify hands, extracting the central position and calculating the areas of hands, using these elements to build hidden markov model. Finally to form a system which can understand the meaning of user's hand movement.