Abstract: The interactive information capture is an important factor to enhance the immersed feeling and sense of reality caused in VR systems.By tracking the arm's location of the interactive person,more harmonious way of human-computer interaction (HCI) can be provided.To track the arm motion,the paper presents an arm motion estimation algorithm based on fuzzy predictive control theory (FPC-ARM),through which the feature rings from real-time images can be extracted and the arm's position can be predicted,to achieve arm navigation operations.The test result indicates that,under the condition with stable environment illumination,the algorithm can stably produce precise information of arm's location.The work is of significance in theory and valuable in practice to immersing VR systems.