Abstract
Abstract. This work presents an extended photogrammetric pipeline aimed to improve 3D reconstruction results. Standard photogrammetric pipelines can produce noisy 3D data, especially when images are acquired with various sensors featuring different properties. In this paper, we propose an automatic filtering procedure based on some geometric features computed on the sparse point cloud created within the bundle adjustment phase. Bad 3D tie points and outliers are detected and removed, relying on micro and macro-clusters analyses. Clusters are built according to the prevalent dimensionality class (1D, 2D, 3D) assigned to low-entropy points, and corresponding to the main linear, planar o scatter local behaviour of the point cloud. While the macro-clusters analysis removes smallsized clusters and high-entropy points, in the micro-clusters investigation covariance features are used to verify the inner coherence of each point to the assigned class. Results on heritage scenarios are presented and discussed.
Highlights
Data filtering plays a fundamental role in 3D reconstructions as it helps e.g. to reduce the noise produced by acquisition sensors and processing procedures
A filtering step can be applied in different phases of the 3D reconstruction procedure: it can be done on the images, on the sparse point cloud, on the dense point cloud or meshes
The goal of this work is to investigate the effectiveness of covariance features (Chehata et al, 2009; Mallet et al, 2011) in identifying and removing outliers in photogrammetric sparse point clouds obtained within a bundle adjustment procedure
Summary
Data filtering plays a fundamental role in 3D reconstructions as it helps e.g. to reduce the noise produced by acquisition sensors and processing procedures. Noisy results are especially frequent when different sensors and platforms are employed, due to scale and illumination changes or quality of single sources (Figure 1). Many filtering algorithms and methods have been developed and included in the processing pipelines (XianFeng et al, 2017). Filtering is even more critical in heritage 3D reconstruction applications, where objects are generally characterized by frequent surface variations and finely detailed elements. A filtering step can be applied in different phases of the 3D reconstruction procedure: it can be done on the images, on the sparse point cloud, on the dense point cloud or meshes. Most of the developed methods have been devised for filtering meshes. Removing outliers and bad computed 3D points in the raw data level (Bastonero et al, 2014) is more convenient with respect to computational efforts and filtering effects and results
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
