Design And Implementation Of Intelligent Evaluation And Correction Method Of Dance Posture

The advancement of society has seen a dramatic rise in people’s quality of life,prompting the swift growth of dance art.However,the drawbacks of traditional dance teaching methods have been gradually becoming more apparent.In the traditional dance teaching mode,the instructor can only judge and analyze the completion of movements based on their experience.However,this method has increased the teaching pressure of dance teachers,and the teaching resources cannot keep up with the increase in demand.This situation,to a large extent,hinders the speed of the development of dance art and the progress of each dance student in dance projects.At the same time,with the development of science and technology,computer vision is constantly integrated into every field of human society,which can help people live more comfortable.This thesis,in accordance with the software engineering development process,designed and implemented a set of intelligent dance posture assessment and correction systems to address the issues posed by posture assessment,thus enabling dance learners to practice online.Help dance learners to carry out dance practice system scheme according to the specific process of dance teaching research,discussion,analysis and design,and finally abstract out the main functions of the system: dance movement analysis and evaluation and dance movement correction guidance.In this thesis,a method combining shape context algorithm and DTW dynamic normalization is proposed to calculate the similarity matching sequence between the sequence of human posture skeleton nodes,and to complete the quantitative analysis of the accuracy and completion of dance learners’ movements.At the same time,LSTM neural network is introduced,and the network model can realize the auxiliary analysis of dance movements after the training of the dance posture skeleton node data.Divided into several modules,the system is comprised of bone matching,motion analysis,motion correction,and machine learning evaluation.In the application,the specific technical details,python language used,and structure of essential code segments and database tables are presented in order of function.Moreover,to facilitate maintenance and expansion,the system combines C/S mode.In terms of database service,according to the performance requirement analysis and use case design of the project,it is decided to use lightweight My SQL database,while maximizing the feasibility of the system in terms of technology and economy.The design and implementation of this system can enable dance learners to practice dance without the restrictions of the venue,and timely understand their learning status through systematic analysis and feedback,and carry out efficient dance practice.