With the development of human-machine interface technology, natural human-machine interaction is becoming an important trend of information development. The communication without wearing any device can be more intuitive and user-friendly. Because people often use body language to support communication, gesture is naturally a very important means of communication. In order to achieve the two-hand gesture recognition, this study uses Microsoft's Kinect sensor device to get the visual information. Compared with a general camera, Kinect can quickly get the depth information that is less affected by environmental light. First, the convex hull algorithm is used to calculate the number of fingers with palm position to determine the static gestures. And then, the strings corresponding to a sequence of static gestures are analyzed by using some fuzzy matching algorithms. These strategies of identify the continuous two-hand gesture can be used as the basis of human-machine interaction.