The Study Of Stereo Matching Algorithm Based On Video Sequences In Binocular Vision

Vision is the main way to obtain information from the outside world,and the study of how to simulate human vision and then apply to the actual application to solve many problems,it has great importance in medical,aerospace,entertainment and other fields.With the rapid development of the computer,it provides a powerful technical support for the visual simulation,which is a hot subject—computer vision.Computer vision combines the computer and the camera,and the camera is used to carry on the imaging and the computer is used to carry on the perception,recognition and analysis to the digital image,finally realizing cognition understanding to the real scene.Binocular stereo vision is one of the most important issues in computer vision,and the stereo matching is the core of it.The disparity results obtained by stereo matching can reflect the distance of the object in 3D space to a large extent,this makes the stereo vision can obtain the depth information which can't be provided by the plane vision.Therefore,stereo matching plays an important role in the field of stereo vision,and it is of great significance for the study of stereo matching.In addition,although the researchers have made a lot of research on stereo matching and proposed a large number of stereo matching algorithms which achieved good results,but they are mainly for the static images,few of them are for video sequences.However,for dynamic scenes such as video sequences,make the full use of the dynamic information that obtained from the video sequences can be good to improve the stereo matching.So this paper studies and proposes a stereo matching algorithm based on video sequences,which is based on the importance of stereo matching and stereo matching of video sequences.In contrast to existing local approaches,a new computational model is proposed in this paper,which is used to deduce and model the adaptive weight parts,and to improve the accuracy and precision of the optical flow information in video sequences.Finally,through a lot of experiments and analysis,the effectiveness and feasibility of the algorithm is verified.