AUTOMATED REGISTRATION OF BUILDING SCAN WITH BIM THROUGH DETECTION OF CONGRUENT CORNER POINTS

Abstract. Current methods of construction progress monitoring involve manual data collection and processing, which are time-consuming and labor-intensive, with a dominant human presence entailing several flaws such as missing or inaccurate information. Recent research efforts for automated progress monitoring have largely focused on model-based assessment methods that are dependent on a pre-requisite step known as registration which is still performed manually due to numerous challenges. This study proposes a novel automated coarse registration method that utilizes the BIM model as the as-planned model to align it with the corresponding as-built model using their geometrical features. First, it extracts the corner points in both models using their planar features and then identifies the conjugate corner points based on different geometric invariants. Later, the transformations are determined from those conjugate points and the most accurate transformation parameter is finalized in the end. The proposed method is validated on different datasets.