| In recent years,motion capture technology has been widely used in various fields,which not only promote the development of society,but also bring many conveniences to people's lives.The motion capture technology has obviously become a popular topic.In response to the development of motion capture technology,a method of motion capture of flexible object based on points feature is proposed.We discuss two aspects: the realization process of the flexible object motion capture method based on points feature and the practical application of the method.The main research contents are as follows:Firstly,the relevant theoretical background and technology are discussed in this paper,it is mainly divided into binocular stereo vision,the classification and method of motion capture technology,and the realization process of three-dimensional reconstruction based on binocular stereo vision is elaborated in detail,which lays a foundation for subsequent research on the method of motion capture of flexible objects based on point-like features.Secondly,we discuss the whole technology realization process of the flexible object motion capture method based on point-like feature in detail,the key steps are including image acquisition,camera stereo calibration,stereo correction,extraction and matching of feature points,and three-dimensional reconstruction.In this paper,we use method called DOG to extract the point feature,then match the point features through the search strategy based on SAD algorithm combined with the RANSAC algorithm to eliminate error matching,and lastly reconstruct by three-dimensional reconstruction of space points and the trajectory reconstruction of time series.Finally,we extract and analysis the three-dimensional spatial coordinates,velocity and acceleration data of the feature points,and drive dynamic data as the static picture,so that the static image is displayed in the same motion effect as that of the flexible object in the experiment,thus realizing a simple application of three-dimensional computer animation.The analysis of the experimental results shows that the method of extracting and matching the points feature used in this paper is more accurate than the traditional SIFT algorithm and SURF algorithm.What's more,we can see the ideal three-dimensional spatial position effect diagram of the points feature shows the motion trajectory of the points feature obtained in the MATLAB environment.The results which prove the validity and feasibility of method proposed in this paper are satisfied. |
