The Design And Implementation Of The Modeling And Driving Subsystem For Sports Human

In recent years,the significant improvement of hardware equipment performance has greatly promoted the development of the virtual character model,which is implemented in the fields of virtual reality,augmented reality,three-dimensional vision.The technology of the virtual character model meets the growing needs of the public in visual perception.However,there is still room for improvement in the visual effect of most of the research on the virtual character model,and there is also still room for improvement in the personalized character modeling and animation display.what's more,the vast majority of applications need the support of special hardware equipment,which is not conducive to cost control and application promotion.Therefore,our research group proposed and designed a portable 3D human motion reconstruction system which is based on the mobile phone,and the system is used to complete the task of action analysis.The system uses the client-server structure to form a simple human motion reconstruction system and collects the action video through the mobile phone.Based on the application of virtual character modeling and driving in the field of motion analysis,we design and implement the moving character modeling and driving subsystem.The subsystem is designed with reference to the MVP structure.The presentation layer implements the display of motion reconstruction results.The data layer implements the loading of body data and motion data.The business logic layer is divided into four modules:three-dimensional reconstruction data processing module and personalized characters.The modeling module,the skeletal skin animation driving module and the animation rendering and display module realize that the virtual character and the real moving character are similar in shape and movement after reconstruction.The system incorporates personalized virtual character modeling on the basis of motion recovery and reconstruction.Compared with traditional motion recovery,the subsystem reconstruction results have a better sense of substitution and observation value.At the same time,the subsystem is based on a graphics engine for the development of skeletal skin animation,which has better visualization effects than traditional methods,and provides a new solution for modeling and driving virtual characters.In order to realize the strong compatibility of the subsystem to various types of data,this paper proposes a data processing algorithm for motion based on Unity3D Engine(DPAM-UE),which is a motion data analysis algorithm based on Unity3D engine.The algorithm consists of a large number of three-dimensional mathematical operations and is divided into three main parts:conversion between motion data,unified description of the coordinate system,and calculation of the global coordinate system data.The algorithm can analyze the motion data in the form of Euler angle,rotation matrix,axis angle pair,and quaternion,etc.,which reduces the data input constraints of the system.At the same time,the algorithm can adapt to a variety of descriptive coordinate systems:it has analytical capabilities for the data in the local coordinate system and the global coordinate system,and has processing capabilities for the data in the left-hand Cartesian coordinate system and the data in the right-hand Cartesian coordinate system,making the subsystem is compatible with data from different motion capture devices and different deep learning models.In terms of structure,this paper first describes related research technology.Then,the requirements of the system are analyzed,the outline design is carried out,and the detailed design and implementation are completed.Thirdly,the paper carries on the function test and the result analysis.Finally,the paper summarizes the work of the algorithm and system and looks forward to future work.