Semi-automatic orientation of images with respect to a point cloud system

Abstract

This paper presents the results of a master thesis in which it was tried to orient a set of images of an object to a point cloud of the same object. As test object the “Semper Sternwarte” in Zürich was used. As data sets a dense point cloud from a laser scanner and a sparse point cloud obtained by photogrammetric means, as well as the orientation of the images were used. Precisely orientated images with respect to a point cloud can be used to incorporate an edge-constrained triangulation techniques, blunder and outlier detection, which leads to an overall better representation of a 3D model. This paper looks at a specific semi-automated work flow using self-programmed tools and tries to determine whether the work flow is suitable for this task or not.