Application Research Of Digital Media Technology In Sports Dance Teaching

With the rapid development of Virtual Reality (VR) technology, simulation teaching system for sports dance obtains more and more attention by scholars working in sports teaching and research and development of computer technology. VR provides teachers and dancers with advanced scientific training methods and widens training scope. Meanwhile, sports dance simulation can stimulate the development of computer science.This thesis mainly studies how to apply VR to rehearse sports dance and constructs a simulation system for sports dance rehearsal. This system can provide technology parameters of sports dance, such as arrangement of parade formation, reasonableness of mechanics, and coordination of acoustic and dance. It is a powerful tool for practitioners. This thesis is arranged as follows. First, we review the current research of sports simulation system based on VR. Then we discuss the virtual human animation, motion edition, and coordination of acoustic and dance to simulate the design and rehearsal of sports dance. At last, we corporate these three aspects into sports dance designing system.In the respect of human animation production, the frames are labeled manually in conventional algorithm, which has three weaknesses:high workload, hard to capture motion, low precision. In this thesis, we combine linear interpolation with four-element spherical interpolation to interpolate the captured motion information. This resolves the problem of Animation frame skipping by using one individual interpolation and makes the animation natural and smooth. The original animation always has many problems. To let amelioration focus on action subject, we have done precise data analysis on every motion of rigid body, solved the problem of shift and rotation by using key-frame interpolation and four element spherical interpolation. The experiments shows that both coach and students can observe their real-time motions and designed motions also can be displayed perfectly. The proposed system turns the training method from experience-based to simulation-based.We request that two motion fragments must be similar on start and end postures in editing dancer’s motions. This is operable in practice. The proposed method obtains human motion parameters in real life, then uses methods of hybrid displacement mapping, affine transformation, and motion image to produce a new motion series from motion fragments selected by user. It works well in experiment.In regard to the coordination of acoustic and dance, we first classify the musics by affection after analyzing the relation between motion and music and then construct a motion database based on affection style. At last, we investigate the technology of showing motion and music synchronously in computer and the feasible impletion method of the model synchronizing audio and video.