Research On Intelligent Character Control And Human-Computer Interaction In Virtual Learning Environment

With the continuous development of virtual reality technology and the pressing need of objective application,virtual character technology has gradually become a research hotspot of virtual reality.As long as human behaviors such as actions and perceptions need to be simulated,virtual characters are indispensable,so virtual character technology is crucial.Through the traditional ways of human-computer interaction,it is difficult for people to fully immerse themselves in a virtual environment.Introducing somatosensory technology into the virtual environment not only makes the human-computer interaction become natural,but also enhances the realism of the virtual environment.In this thesis,virtual reality technology,human-computer interaction technology and traditional teaching method are combined to build a human-computer interaction system oriented to virtual learning environment,and intelligent character modeling and human-computer interaction technologies in the virtual learning environment are studied.The main contents in this thesis are as follows:1.A framework for building the human-computer interaction system oriented to virtual learning environment is proposed,on this basis,the thesis analyzes the key technologies of virtual reality system.The intelligent character 3D model and behavioral model with visual perception,auditory perception and behavioral decision are built.Besides,the implementation procedure and the used methods are introduced.Studying on path planning technology,the grid graph is used to represent the virtual environment,and the path planning of the intelligent character based on A* algorithm is implemented in the virtual learning environment.2.The human-computer interactions based on virtual character and the operational principles of Kinect are studied.The skeletal joint point information of human body is acquired by Kinect,and then the acquired skeletal joint points are bound with the skeletal joint points of virtual character to realize virtual avatar interaction.The virtual avatar's postures are recognized by calculating the distances and angles among the skeletal joint points,on this basis,realizing the posture interaction based on the virtual avatar.Kinect is used to collect and process speech information,and then speech recognition is performed.Next the speech recognition results are converted into commands to control the actions of the intelligent character,enabling speech interaction based on the intelligent character.3.Developing a human-computer interaction system oriented to virtual learning environment,we realize the functions for the system including scene roaming,PPT presentation,book browsing and inverted pendulum experiment operation.Finally,in the virtual learning environment the main system functions and the system performance are tested.The results verify the effectiveness of the system.