3D scene scanning automated method from the motion image

摘要

System that generates automatically (57) a three-dimensional model that is completely texturing of the target from the summary motion image is disclosed. Using the operators of the salient features, the system keeps track of the salient features of the motion image based on the salient features that are detected. Are used to construct the feature blocks containing prominent features to be tracked, each feature block, feature tracking map as input to the camera motion estimation process is controlled to provide a solution to the fluoroscopy camera motion estimation process dense points used to present and that generates a mesh model is extracted. Finally, three-dimensional model mesh model is texturing the motion image, has been texturing completely by it is generated.