3D Object Reconstruction And Recognition Method Based On Visible Hull

In the mechanical processing, virtual reality,3D communication and medicalapplication,3D reconstruction and recognition have a broad prospect of application.3Dreconstruction and recognition technology are also important parts of machine vision. Thispaper puts forward a kind of3D object reconstruction and recognition method based onvisible hull, this method can realize the automatic reconstruction and recognition ofmachine vision, and it has a strong application value.3D reconstruction method rebuilds objects through the relationship between cameraimaging coordinate system and objects, so the camera calibration is the basis ofreconstruction and recognition system. This paper proposes a camera calibration method.Comparing with the traditional calibration method of complicated operation, this selfcalibration method does not rely on the certain reference system to get camera parameters,only through the relationship between the camera and the movement of objects. After usingthe threshold segmentation and region growing algorithm edge information processing toget the target contour, we can create3D object reconstruction with the visual hull method.The3D object reconstruction method uses seed body of grain growth way to complete theadaptive visible hull. World coordinate system can make the object in translation, rotationand direction change with constant qualitative and robustness, and it just needs to know theobjects surface coordinates. Then we can use the world coordinates similarity matchingmethod to identify the object.After finishing the camera calibration, image preprocessing, visible hull3D objectreconstruction, and the world coordinate similarity matching object recognition methodthis paper puts forward a machine recognition system, the system can realizereconstruction and recognition of simple3D objects. Related experiments and systemoperation show that the method can realize the reconstruction and recognition of machinevision, and have high precision and robustness.