Capturing And Application Of Human Motion Information Based On Kinect

Capturing, tracking and recording the human body information in three dimension by sensor equipment is a hot research areas in current year. This kind of technology has played an important role in film production, medical research, military and intelligence control fields. For example, motion capture device is used to record human movement for movie special effects, or used in building human model for medical research. However, the most used of motion capture system is based on optical passive capture equipment, which is expensive, complex installation and cumbersome in using process.As a new type of motion sensing equipment, K inect has been successfully used in medical surgery, elder care and motion sensing game. In order to build an portable capture system and fulfill the analysis of motion parameter automatically, the human information were captured and analysed from three kinds of data flow caught by K inect in this dissertation. Kinect has been gave full play to its strong extensibility and the characteristics of convenient in this research. The specific content is as follows:Firstly, in order to authenticate the validity of angular velocity and angular acceleration parameters captured by K inect,three kind of movement were performed in experiment. The three dimensional motion information was automatic analysed automatically by K inect. O n the other hand, all data were compared through Vicon system to comparison analysis. The results of experiment show that all Pearson correlation coefficient of the data of are greater than 0.7 with significant correlation, which suggests that K inect can be used as an analysis tool for kinematic variability of shoulder and trunk motion, and could help obtain low back information regarding movement patterns and strategies under various movement tasks.Secondly, I put forward to a method that calculating three dimensional center of mass of human body using K inect, and the data results were compared with the motion capture system. The conclusion shows that the application has great potential to be evaluating tool of human body balance ability. At static experiment movement, the mean of the root mean square error are range 1 mm and 3.8 mm,and this parameter are range 3 mm and 7.4 mm in incline status. The value are 7.5mm?6.7mm and 5mm in X?Y?Z axis respective. K inect could reflect the static standing human body center of mass trajectory trend better, and is suitable for the usage of large movement. Overall, K inect could be used in measuring the three-dimensional center of total mass of human body and balance ability evaluation by software programming and algorithms, and this application will be more widely used in medical field and clinic.Thirdly, upper limb rehabilitation training system have been developed based on the Kinect. The rehabilitation interactive games using coordinate conversion, real animation effects, background of fusion, filtering technology and so on. This rehabilitation system can make the interactive participants immersed in a game, enhance participants' interest and improve the training effect, achieving the goal with highly intense, the task-oriented and repeated rehabilitation training.