Rock, Paper and Scissors image identification scheme

In this paper, several image processing techniques are used to achieve recognition of Rock, Paper, and Scissor hand signal identification. First, we extract the hand image from the video sequences. Next, the generalized threshold and skin color are employed to generate a binary image and filter-out interference. Next, the distance between the tips of the two fingers is used to get rid of those obviously error states. However, several Scissors states are difficult to correctly judge. Successively, the segmentation, histogram, and angular criteria methods are used to distinguish the exactly Scissors hand signal. It is the concept that the extended line of the hand (segment CD) shown in Fig. 7 must fall in the included angle (segment AD and BD). Finally, the scheme calculates the conditionals m_BD ≥ m_CD ≥ m_ADand m_AD ≥ m_CD ≥ m_BD which are used to judge whether the Scissors is correct or not. Simulation results demonstrate that our scheme is an effective method for Rock, Paper, and Scissor identification.