Three Dimensional Vessel Tree Reconstruction By Multi-view Stereo Reconstruction

Image-based modeling is a method of modeling which has been paid attention in recent years. It can get the realistic model of a specific object. The 3-D (three dimensional) reconstruction of tree-like objects has important application in medicine, biology and the simulation of natural scene.3-D reconstruction of the arterial tree is significant for providing accurate quantification of the coronary vasculature and stenosis assessment. Thus,3-D reconstruction of tree-like objects based on the two single plane angiograms is significant and valuable. This thesis expatiates the problem of 3-D reconstruction of tree-like objects from images obtained from the rotational C-arm angiography system and discusses the method of reconstructing the branch models of a tree from multi-view stereo reconstruction. During C-arm angiographic imaging, the contrast bolus flowing through blood vessels may sustain a bolus/blood equilibrium state in the scan field of view, which can be considered as quasi-static or stationary state. The C-arm images can be used for vessel tree reconstruction by searching the corresponding points in two projection images followed by two-view stereo reconstruction.Major contents of this dissertation include the following:(1) Introduce the basic mathematic knowledge, basic theories and methods of the multiple view geometry, research the algorithms of estimation of fundamental matrix, and then use random sample consensus (RANSAC) weighted translation algorithm to obtain the fundamental matrix with high accuracy and robustness;(2) Introduce the imaging theory of rotational C-arm angiography system and add the topology constraint to epipolar line constraint to realize the automatic matching of feature points; reconstruct the tree phantom based on the automatic matching of the feature points in projective space and then realize Euclidean reconstruction with a cube phantom calibration; finally, use Matlab to show the 3-D sketch of the tree phantom and perform 3-D tree measurements (tree segment lengths) and compare with the reconstructed results to verify the effectiveness of the method in this paper.