TU Darmstadt / ULB / TUbiblio

Removing the Example from Example-based Photometric Stereo

Ackermann, Jens and Ritz, Martin and Stork, André and Goesele, Michael (2012):
Removing the Example from Example-based Photometric Stereo.
Springer, Berlin, Heidelberg, New York, In: Trends and Topics in Computer Vision, In: Lecture Notes in Computer Science (LNCS); 6554, DOI: 10.1007/978-3-642-35740-4₁₆, [Conference or Workshop Item]

Abstract

We introduce an example-based photometric stereo approach that does not require explicit reference objects. Instead, we use a robust multi-view stereo technique to create a partial reconstruction of the scene which serves as sceneintrinsic reference geometry. Similar to the standard approach, we then transfer normals from reconstructed to unreconstructed regions based on robust photometric matching. In contrast to traditional reference objects, the scene-intrinsic reference geometry is neither noise free nor does it necessarily contain all possible normal directions for given materials.We therefore propose several modifications that allow us to reconstruct high quality normal maps. During integration, we combine both normal and positional information yielding high quality reconstructions. We show results on several datasets including an example based on data solely collected from the Internet.

Item Type: Conference or Workshop Item
Erschienen: 2012
Creators: Ackermann, Jens and Ritz, Martin and Stork, André and Goesele, Michael
Title: Removing the Example from Example-based Photometric Stereo
Language: English
Abstract:

We introduce an example-based photometric stereo approach that does not require explicit reference objects. Instead, we use a robust multi-view stereo technique to create a partial reconstruction of the scene which serves as sceneintrinsic reference geometry. Similar to the standard approach, we then transfer normals from reconstructed to unreconstructed regions based on robust photometric matching. In contrast to traditional reference objects, the scene-intrinsic reference geometry is neither noise free nor does it necessarily contain all possible normal directions for given materials.We therefore propose several modifications that allow us to reconstruct high quality normal maps. During integration, we combine both normal and positional information yielding high quality reconstructions. We show results on several datasets including an example based on data solely collected from the Internet.

Series Name: Lecture Notes in Computer Science (LNCS); 6554
Publisher: Springer, Berlin, Heidelberg, New York
Uncontrolled Keywords: Business Field: Virtual engineering, Business Field: Visual decision support, Research Area: Generalized digital documents, 3D Scene reconstruction, Photometry, Multi-view stereo, Computer vision, Surface reconstruction, 3D Scanning, 3D Reconstruction
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Interactive Graphics Systems
Event Title: Trends and Topics in Computer Vision
Date Deposited: 12 Nov 2018 11:16
DOI: 10.1007/978-3-642-35740-4₁₆
Export:

Optionen (nur für Redakteure)

View Item View Item