Research Of Virtual Environment Simulation And Human-computer Interaction Based On Unity3D

Virtual reality technology,as an emerging technology that involves multiple fields has been wide concerned.Especially in recent years,virtual reality technology broke out suddenly and started to take steps in the people's life.This technology is being applied in all walks of life gradually.How to construct a virtual environment more realistically and precisely,and to interact with users effectively have been become one of the key contents of researchers.Previously,most of the virtual reality applications used tools such as UE4,but its development process was complicated and the external capabilities were poor.With the release of Unity3 D,its efficient development process and rich component have make a great advantage in the development of virtual reality.The purpose of this thesis is to simulate a more exquisite and natural virtual environment by the Unity3 D engine and to complete the interface between reality and virtual reality through human-computer interaction.The work and contributions of this thesis mainly concentrated on the following aspects:(1)In the global light extraction method based on light probes,the layout of the light probes in the scene are optimized,and the time of illumination pre-acquisition is shortened effectively while ensuring the rendering effect.(2)This thesis realized the motion simulation of virtual characters through human motion data based on the key frame editing method.In this method,we collect the position data of the joints of the real human body through the equipment and then calculate the rotation angle of each joint.In order to ensure the rationality of the motion information,we restrict the motion angle of each joint of the human body and complete the motion simulation of the virtual character by the interpolation between the key frames.(3)This thesis analyzed the existing human-computer interaction method and proposed a human-computer interaction method based on monocular camera.In this method,we divide the 3D skeleton extraction of human body into two parts:2D skeleton extraction and non-parametric 3D skeleton matching.In the non-parametric3 D skeleton matching method,we add the continuity principle of human motion,which guarantees real-time performance on the basis of correctness.Furthermore,we propose a human-computer interaction method based on hand motion informationthrough skeleton information extracted from monocular camera.(4)On the basis of above research results,the virtual robot motion system based on monocular camera,children's educational system and deep space detection system are designed and implemented by the Unity3 D engine.