Design and Implementation of Teaching System Based on RGB-D Human-Computer Interaction.

The traditional teaching model has shortcomings such as single presentation method and difficult storage of writing on the blackboard, The emergence of multimedia teaching will change the form of chalk and blackboard in traditional teaching and improve teaching efficiency. However, there are also defects such as relying on input device control, which affects teaching quality. The control method based on human-computer interaction can solve this drawback and improve the interaction between teachers and students. Therefore, this paper designs a teaching system based on RGB-D human-computer interaction. This system uses Kinect to obtain depth images to identify the skeleton model of the human body. By analyzing the needs of the teaching system, designing corresponding control actions and acquiring three-dimensional data of key joint points for these actions, Analyze this data to get the main feature information and auxiliary judgment information of the movement, Design a control action recognition algorithm based on this information, Finally, comprehensive control requirements and control actions are designed to design a teaching control system to realize auxiliary control of common teaching functions. After testing, the correct rate of the system's various action recognition algorithms is about 95%, and the real-time, stability and other indicators meet the requirements of teaching control, and the system performance is good.