Autonomous image background removal for accurate and efficient close-range photogrammetry

Abstract Close-range photogrammetry can be used to reconstruct dense point clouds of an object with very high surface coverage, making it useful for manufacturing metrology tasks such as part inspection and validation. However, compared competing techniques, data processing times slow. In this paper we present a method autonomously remove the background from images within photogrammetric dataset. We show that using masked directly in reconstruction results much lower times, memory utilisation. Furthermore, density on is increased while number superfluous points reduced. Finally, set are tactile coordinate measurements. Reconstructions removed shown have standard deviation mesh distance up 30 µm than if not removed. This improvement likely due static background, relative rotation stage, causing triangulation errors when detected matched data. The proposed approach robust over several example artefacts can, therefore, implemented improve measurement efficiency systems.