Research On 3D Face Reconstruction And Recognition Method Based On Binocular Stereo Vision

Nowadays,the continuous updating of social science and technology,binocular machine vision in the field of artificial intelligence has once again become the focus of people's research.The advantages of machine vision technology are low cost,non-contact,easy to implement and so on,so it is widely used in road detection.And navigation,product defect detection,medical equipment,target location and tracking.Object recognition is a basic research in the field of artificial intelligence.An important branch of biometric recognition is face recognition.It is widely used in the fields of video surveillance,identity authentication,human-computer interaction and security anti-terrorism.Changes in lighting,expressions,facial gestures,etc.can adversely affect face recognition.3D data is robust to illumination and attitude,and it is of great significance to improve the recognition accuracy of face recognition in actual scenes.Therefore,it is very important to create realistic 3D face models.Based on the above background,this paper uses a binocular machine vision system to identify three-dimensional faces.Starting from the face image taken by the binocular camera,the two face images obtained by the left and right cameras are reconstructed three-dimensionally and the reconstructed three-dimensional face is recognized.The binocular stereo vision bionic human eye detects and locates the object in space,and calculates the spatial position coordinates of the point on the object according to the image of two different viewpoints and the similar triangle formed by the points on the object in the space scene.The most important part of 3D reconstruction is to use stereo matching to obtain parallax,and to derive the depth of the target according to the triangle similarity principle,that is,the distance from the target to the human eye.Based on SURF matching algorithm and SIFT matching algorithm,this paper proposes a new stereo matching algorithm based on thisalgorithm.By obtaining the difference of the abscissa of the matching points of the images acquired by the left and right cameras,the matching is calculated according to the coordinate transformation theory.The position coordinates in the space,the 3D point cloud obtained by the improved stereo matching algorithm is used for 3D face modeling.3D face reconstruction and recognition based on binocular stereo vision,including camera calibration,image preprocessing,stereo matching and 3D reconstruction and recognition.In this paper,the binocular stereo vision device is used to build the experimental platform.Two cameras are used to obtain the face image pair.The internal parameters and relative position of the camera are calibrated by Zhang Zhengyou's checkerboard calibration method.The image preprocessing after calibration correction is proposed.The new stereo matching algorithm acquires the disparity map for the face image,and finally obtains the three-dimensional coordinate information of the face,thus verifying the feasibility of the method in recognition.