A Platform Ot Automatically Generating Labanotation Based On Extreme Learning Machine

With the development of society,more and more traditional dynamic cultures such as folk dances,martial arts,and operas cannot be effectively protected.How to effectively pass them down is a problem that needs to be resolved.As a well-recognized scientific action analysis and recording system,Laban Spectrum is applied to the recording of traditional dynamic culture to have the advantage of being easy to inherit.However,it is time-consuming and labor-intensive to manually record the Laban dance spectrum.Therefore,how to use computer technology to automatically generate Laban dance spectrum has become an important research topic.The main work of this dissertation is to propose the Laban dance spectrum algorithm based on extreme learning.By analyzing the motion capture data,on the premise of segmenting the motion,each basic motion is identified and converted into the symbol of the Laban dance spectrum.On the basis of this algorithm,an automatic dance spectrum generation platform is realized.In addition,the design of the digital record standard of Laban dance was completed in light of the shortcomings of the current file format for the Laban dance spectrum file.In order to enrich the functions of the platform,the platform has realized the function of transforming dance music into animation.The main work and innovations are as follows:(1)In order to improve the accuracy of dance spectrum generation,this paper proposes a dance spectrum generation algorithm based on extreme learning.First,the acquired human motion capture data is converted into coordinates in the world coordinate system,and then the segmentation action segment data is obtained through segmentation.Finally,the extreme learning algorithm is used to generate the dance spectrum symbols for these segmented action clip data,display and record them,and compare them with the dynamic programming and the support vector machine in terms of accuracy and efficiency.(2)In order to record the Laban dance spectrum more conveniently,this article has designed a standard format for digital recording of the Laban dance spectrum.At present,there are two ways to record the action system.The first one is through video recording,which takes up a large amount of memory;the second is the way that pictures are added to the text to record,which is not intuitive enough.The record format of the Laban dance spectrum is in the form of pictures.The advantages of this method are relatively simple,but there are also disadvantages that the storage format is inflexible and cannot be understood by the computer.This article has completed the development of the digital record standard of Laban dance spectrum for this situation,and it is implemented in the automatic generation of Laban dance spectru(?)latform.(3)In order to better carry out traditional cultural protection work,based on the extreme learning algorithm,this article has realized the automatic generation platform of Laban dance spectrum.This platform also includes the ability to convert the Laban dance spectrum to animation.In the selection of the motion capture system,the optical motion capture system was chosen in this paper,mainly considering its simplicity and ease of operation.First,establish a three-dimensional motion model of the human body and then collect the human motion data,and save the human motion data as a BVH(Bio-vision Hierarchical)format file.Through the conversion and analysis of BVH data,the automatic generation platform of Laban dance spectrum is realized.In order to better carry out related work,the platform also implements the function of converting the Laban dance spectrum into an animation,so the platform realizes a closed loop from dynamic actions to static dance spectrum to dynamic actions.This is the first one The action is converted into a Laban dance spectrum and you can convert the Laban dance spectrum into a platform for action.