| The human body as a social environment,contains a variety of interactive behavior information,the study of body posture behavior profound social and practical value.In particular,with the promotion of three-dimensional scanning technology,increasing the number of three-dimensional model of human surge,and,in the field of motion simulation,film and television entertainment,advanced human-computer interaction been more and more widely used.Therefore,how efficient and accurate three-dimensional anatomical information extraction and analysis to achieve the reuse model has become an important research content.Develop effective three-dimensional structure of the human body feature extraction methods have become more in-depth study of human behavior on the basic of work.By structural features extracted achieve body posture matching model behavior can be better understood,and highly practical in terms of motion analysis,human model retrieval,human-computer interaction.Extraction algorithm to extract the presence of the traditional two-dimensional skeletal structure skeleton features not accurate enough,this paper presents the method of human gesture recognition based on the dynamic path optimization skeleton model extraction,,to achieve an efficient,accurate identification body posture behavior,human body model after compression,retrieval applications,simulation and other areas of the foundation.The paper important study contains:1.In order to efficiently and accurately identify the body posture behavior,First capture device through joint-point calibration and operation,obtain real-time three-dimensional body movements and gestures sequence;2.Secondly,according to the space andtime seriesdata,the introduction ICP method to achieve the test model similarity calculation sample data,the establishment of the operation sequence similarity matrix,whereby dynamically determine optimal matching path;3.Eventually,by calculating the overall similarity of sequences of actions to achieve the body posture of behavior analysis.A series of experimental result sverify the stability and efficiency of the method. |