The Research And Application Of Natural Human-computer Interaction Key Technology

With the development of natural human-computer interaction technology, the interaction between people and computer has been gradually transferred to user-centered. Visual-based natural human-computer interaction technology is more suitable for popularization application due to its high degree of freedom, low cost, strong commonality, etc. This paper research work is about three aspects:visual gesture recognition, limb motion recognition and virtual interactive applications.(1) Visual Gesture Recognition Technology Research. Based on skin-color model, a kind of multi-sequences background model is proposed, combing with skin-color block tracking and geometric shape estimating, to extract gesture images. To compute gesture parameters, gesture geometric characteristics are analyzed. To distinguish and recognize different hand gestures, decision tree classification method is adopted. Experimental results demonstrate that the recognition algorithm is simple and effective, small amount of calculation. The recognition feedback time is less than 0.1 seconds. The recognition accuracy rate of the predefined 6 kinds of natural gesture types reaches 94%under the general background and indoor daytime well-lit environment. The gesture recognition system can implement the synchronized multimedia interactive applications effectively.(2) Limb Motion Recognition Technology Research. In this paper, a new posture sequence finite-state machine (FSM) method is proposed. Firstly, a user-centric limb joint coordinate system is setup to transform the skeletal data of the user motions from the device space to the user space, and a three-dimensional grid model is built to eliminate the users’ individual differences. Secondly, the joint motion sequences of predefined motions are sampled and analyzed using the limb joint feature vectors, which are defined in the limb joint coordinate system, with spatial motion vectors, motion time intervals, and spatial distances of joints being employed to depict the limb motion features. Finally, the limb motion trajectory is represented by the joint movement regular expressions, and the posture sequence FSM is constructed to recognize the predefined motions online. Experimental results on 17 kinds of predefined motions show that the proposed method has the merits of high recognition rate more than 94%, low feedback time less than 0.1 seconds, good scalability and versatility, and is suitable for natural human-computer interaction applications.(3) Natural Human-Computer Interaction System Implementation. In this paper, three natural human-computer interaction applications: Wisdom Tourism Virtual Interaction Experience System, Intelligent Housekeeper Robot SHRobot, Virtual Assembly Interaction System, are designed and implemented based on the key visual gesture and limb motion recognition method. Meanwhile, a good versatility and adaptability of natural human-computer interaction system framework is proposed. Direct and humanized interaction functions for interactive applications are provided by natural human computer interaction mode. Interactive experience and immersive feeling are increased according to the all-round, multi-angle real-time virtual interactions with specific application scenario.