INVESTIGATING THE EFFECT OF FOCUS STACKING ON SFM-MVS ALGORITHMS

Abstract. The Depth of Field (DoF) is a vital factor in photogrammetric applications. Its effect is in most cases pretty obvious especially when capturing small artefacts. It is very important to observe its behaviour as it affects the ability to capture all the details of an object. Focus stacking is a technique in computational photography, in which a set of images focused on different planes with limited DoF are combined in order to considerably extend the DoF. Today, there is a number of focus stacking methods that can be applied in order to produce a full-focus image. In this paper, we investigate the application and effects of focus stacking on SfM-MVS 3D reconstruction. Specifically, our experiment involves the 3D reconstruction of a selected artefact using both traditional all-focus photography and focus stacking. The artefact has already been digitised with a high accuracy and resolution structured light 3D scanner, and that 3D model served as the reference model, with which SfM models were compared. We discuss on these fist results and present some preliminary assessment on the application of focus stacking for the SfM-MVS-based 3D reconstruction.