Fast and Robust Semi-Automatic Registration of Photographs to 3D Geometry

Abstract

We present a simple, fast and robust technique for semi-automatic 2D-3D registration capable to align a large set of unordered images to a massive point cloud with minimal human effort. Our method converts the hard to solve image-to-geometry registration problem in a Structure-from-Motion (SfM) plus a 3D-3D registration problem. We exploit a SfM framework that, starting just from the unordered image collection, computes an estimate of camera parameters and a sparse 3D geometry deriving from matched image features. We then coarsely register this model to the given 3D geometry by estimating a global scale and absolute orientation using minimal manual intervention. A specialized sparse bundle adjustment (SBA) step, exploiting the correspondence between the model deriving from image features and the fine input 3D geometry, is then used to refine intrinsic and extrinsic parameters of each camera. Output data is suitable for photo blending frameworks to produce seamless colored models. The effectiveness of the method is demonstrated on a series of real-world 3D/2D Cultural Heritage datasets.