Font Size: a A A

Taiji Boxing Assist Teaching And Evaluation Method Based On Whole Body Motion Capture

Posted on:2019-08-25Degree:MasterType:Thesis
Country:ChinaCandidate:Z XuFull Text:PDF
GTID:2428330542994590Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Tai Chi is a kind of fitness activity that is popular among the broad masses of the people.It is one of the ways people cultivate their temperament and cultivate their sentiments.Through the use of motion capture systems and data gloves to capture the human movement posture,real-time guidance Tai Chi,to promote Tai Chi,Tai Chi culture has an important significance.This article proposes Tai Chi assisted teaching and evaluation methods that combine the two-way natural interaction of Tai Chi and human body attitude capture,providing a new teaching method for Tai Chi learning enthusiasts.The main work of this article is as follows:(1)Aimed at the problem of real-time human body motion capture and analysis based on the level of the constraints and rigid body mechanics of human body skeleton model,use the inertial motion capture system and data gloves to acquire human posture data,and use 3D surface patch model to construct 3D human body model,real-time dynamics Reconstruct the human three-dimensional skeleton model.(2)Aimed at the problem of the common gestures in Tai Chi,analyze the common gestures in Tai Chi,use data gloves to collect a large number of Tai Chi gesture data,establish a BP network model based on PSO improvement,and put forward the recognition algorithm of Tai Chi in boxing gestures.(3)For Tai Chi overall and local movement trend forecasting and identification problems,analyzing the characteristics of Tai Chi moves,the active conditions and spatial characteristics of major joints in Tai Chi movements,and proposes constructing a Tai Chi movement original language library;analyzing human movements and human body Based on the distribution characteristics of joint sensor,a similarity pyramid model of Tai Chi movement was proposed;Analyzing the whole body sensor position data,based on the SVM's overall motion recognition;analyzing the local joint sensor data form the local joint vector,based on the CCA local joint vector measurement,predict the joint movement trend,and give reasonable advice on Tai Chi movement.The experimental results show that the human body movements and hand postures can be accurately identified in real time,and the action database can be built off-line,and the local joint movements can be analyzed in real-time online.The guidance of reasonable movement guidance can be given in time,which can effectively guide the Tai Chi teaching.Promote the spread and development of Tai Chi.
Keywords/Search Tags:Tai Chi, full body motion capture, artificial neural network, SVM, CCA
PDF Full Text Request
Related items