The Research And Application Of Motion Trajectory Recognition In Stage Performance

Today,the rapid development of computer technology brings forth the related technology and the progress of the industry.People gradually raise awareness of human-computer interaction,to make it easier to interact with the computer,and make the computer more intelligent and serve human beings efficiently.As the prevalence of the film and television product,body feeling game,rehabilitation,medical,sports teaching and stage performance,many researchers have been attracted to study its main technology of action recognition.In the process of action recognition,we first to acquire the data information,while the captured human movement sequence has a big volume of data in both time and space dimension,leading to existing redundant data between frames.In addition,the randomness of human movement brings enormous challenge for the action recognition.Therefore,how to effectively express the movement,extract the characteristics of human body and apply effective method to recognize the human movement have become the critical problems.This paper deeply study the theory and method of human action recognition from the method of extracting and recognizing the characteristics of human action.With the consideration of reducing the reconstruction error and improving character generalization,this paper analyzes and builds models of minimization of error reconstruction and maximization of Character generalization,and finally utilizes the Artificial Bee Colony Algorithm to optimize the calculations of these models.It proposes the method of key-frame extraction for captured movement data,based on the Artificial Bee Colony Algorithm,and this method has achieved the way of extracting the key characteristics of multiple motion captured data effectively.Furthermore,this paper implements simulation based on the captured key-frame and DTW Algorithm to recognize motion and verify the effectiveness of algorithm.In the end of this paper,we combine the key-frame extraction and action recognition algorithm with the stage background switch,and utilize Microsoft Kinect to build the stage background switch system based on motion recognition,and design the corresponding software system based on Unity,and finally utilize the proposed algorithm to achieve the switch between motion recognition and stage background.