Youtor
Introduction to Photogrammetry and Points2Surf ECCV 2020
>> YOUR LINK HERE: ___ http://youtube.com/watch?v=-F7lAin_pMI
This lecture is an introduction to photogrammetry. It covers geometry representations, conversion algorithms and evaluation methods. We explain methods that convert between images, point clouds, explicit surfaces and implicit and surfaces. Covered algorithms include Structure-From-Motion, Poisson Surface Reconstruction and Marching Cubes. While we keep the focus on the typical photogrammetry pipeline, we also present other methods, i.e., our own research on surface reconstruction from point clouds (Points2Surf, ECCV 2020). We conclude the lecture with an overview of common evaluation metrics like F1-score and Chamfer distance. • Lecture slides: https://vapor.cg.tuwien.ac.at/index.p... • • 00:00 Intro - Part 1 Introduction to Photogrammetry • 02:56 Motivation - KinectFusion • 05:11 Geometry Representations • 13:22 Geometry Representations - Point Clouds • 18:21 Geometry Representations - Explicit Surface • 20:41 Geometry Representations - Implicit Surface • 25:37 Methods – Scanning Pipeline • 26:57 Methods – Structure From Motion (SFM) • 34:13 Methods – Multi-View Stereo (MVS) • 41:20 Methods – Poisson Surface Reconstruction • 45:32 Methods - Marching Cubes • 50:07 Methods - Points2Surf • 58:31 Methods – AtlasNet • 01:00:18 Methods – Differentiable Rendering • 01:04:03 Reconstruction Methods Evaluation • 01:24:22 Part 2 - Points2Surf Code (https://github.com/ErlerPhilipp/point...) • • [Chapters were auto-generated using our proprietary software - contact us if you are interested in access to the software] • The reconstruction focuses on the author's paper: • Points2Surf: Learning Implicit Surfaces from Point Cloud Patches (ECCV2020) • Paper: https://arxiv.org/abs/2007.10453 • Git: https://github.com/ErlerPhilipp/point... • • Presenters BIO: • Philipp Erler is a PhD student in the rendering and modeling group (headed by Prof. Michael Wimmer) at TU Wien. His area of research is surface reconstruction using deep learning. The cooperation with Prof. Niloy Mitra (UCL / Adobe) and Dr. Paul Guerrero (Adobe) resulted in his first publication, Points2Surf. He also assists in teaching and reviewing papers. • His homepage: https://philipperler.net/ • • Talk references: • Research group Youtube channel: / @cgtuwien • Marching Cubes variants: • Classic: http://fab.cba.mit.edu/classes/S62.12... • Available as a stand-alone Python package, PyMCubes: https://github.com/pmneila/PyMCubes • Lewiner variant with topological guarantees: https://www.ks.uiuc.edu/Research/vmd/... • Available in skimage (classic version is also available): https://scikit-image.org/docs/dev/aut... • I used this one in Points2Surf • Surface extraction used in Poisson Rec is a modification of an Octree MC. The paper refers to these 3 papers: • WILHELMS J., GELDER A. V.: Octrees for faster isosurface generation. TOG 11 (1992), 201–227. • SHEKHAR R., FAYYAD E., YAGEL R., CORNHILL J.: Octree-based decimation of marching cubes surfaces. In IEEE Visualization (1996), 335–342. • WESTERMANN R., KOBBELT L., ERTL T.: Real-time exploration of regular volume data by adaptive reconstruction of iso-surfaces. The Visual Computer 15 (1999), 100–111. • Unconstrained isosurface extraction on arbitrary octrees • http://hhoppe.com/proj/unconstrainediso/ • This one looks like a good choice for Octree-SDFs but I haven't tried it yet. • Accurate Isosurface Interpolation with Hermite Data (resolving ambiguous cases of classic marching cubes) • https://github.com/mkazhdan/IsoSurfac... • An implementation of Dual Marching Cubes (DMC) for already sampled data: • https://github.com/alexdevteam/dmc • ------------------------- • Find us at: • Newsletter for updates about more events ➜ http://eepurl.com/gJ1t-D • Sub-reddit for discussions ➜ / 2d3dai • Discord server for, well, discord ➜ / discord • Blog ➜ https://2d3d.ai • AI consultancy Abelians ➜ https://abelians.com/
#############################
