Research On Camera Control And Programmino For Virtual Human Movement

With the continuous improvement of people’s living standards, there is a growingfavor of a realistic three-dimensional visual experience. Any three-dimensional visualexperience inevitably depends on the position, orientation and movement of theviewpoint. So control a virtual camera in three-dimensional virtual environment is avery important research topic. However, cameral control algorithms about the virtualhuman animation and crowd animation have rarely been studied at home and abroad. Inthis thesis, we focused on the motion capture data, implemented two camera pathplanning algorithms in the virtual sports scene.Firstly, this paper introduces the camera path planning background and researchstatus. And the skeletal animation, methods of angular parameterization and thetechnology of motion capture are introduced, at the same time the Mocap data structureis described, then we analysis the virtual human motion data and give the definition ofvirtual human motion swept area. Combined with the movement significantly of thevirtual human, we proposed a camera path planning algorithm and smoothed the pointof view of each frame with global interpolation algorithm. The result shows that thealgorithm can plan for a smooth camera path to observe the virtual human movement.A depth study of the establishment of the Kalman filter equations and recursiveprocess is taken. And we use the Kalman filter and a virtual human movement sweptarea to establish a mapping between virtual human motion parameters and cameraparameters, then we defined a suitable description of the scene. And a real-time cameraplanning algorithm for virtual human movement is proposed. The results shows that thealgorithm can achieve real-time observation of the virtual human movement and canshow the details of the movement of the virtual human.