Tinto: Multisensor Benchmark for 3D Hyperspectral Point Cloud Segmentation in the Geosciences

DOI

The increasing use of deep learning techniques has reduced interpretation time and, ideally, reduced interpreter bias by automatically deriving geological maps from digital outcrop models. However, accurate validation of these automated mapping approaches is a significant challenge due to the subjective nature of geological mapping and the difficulty in collecting quantitative validation data. Additionally, many state-of-the-art deep learning methods are limited to 2D image data, which is insufficient for 3D digital outcrops, such as hyperclouds. To address these challenges, we present Tinto, a multi-sensor benchmark digital outcrop dataset designed to facilitate the development and validation of deep learning approaches for geological mapping, especially for non-structured 3D data like point clouds. Tinto comprises two complementary sets: 1) a real digital outcrop model from Corta Atalaya (Spain), with spectral attributes and ground-truth data, and 2) a synthetic twin that uses latent features in the original datasets to reconstruct realistic spectral data (including sensor noise and processing artifacts) from the ground-truth. The point cloud is dense and contains  3,242,964 labeled points. We used these datasets to explore the abilities of different deep learning approaches for automated geological mapping. By making Tinto publicly available, we hope to foster the development and adaptation of new deep learning tools for 3D applications in Earth sciences.

This research received funding from the Initiative and Networking Fund (INF) of the Hermann von Helmholtz Association of German Research Centres in the framework of the Helmholtz Imaging Platform under grant agreement No ZT-I-PF-4-021.

Identifier
DOI https://doi.org/10.14278/rodare.2256
Related Identifier https://www.hzdr.de/publications/Publ-36833
Related Identifier https://doi.org/10.14278/rodare.2255
Related Identifier https://rodare.hzdr.de/communities/hzdr
Related Identifier https://rodare.hzdr.de/communities/rodare
Metadata Access https://rodare.hzdr.de/oai2d?verb=GetRecord&metadataPrefix=oai_datacite&identifier=oai:rodare.hzdr.de:2256
Provenance
Creator Afifi, Ahmed J. M. ORCID logo; Thiele, Samuel Thomas ORCID logo; Rizaldy, Aldino; Lorenz, Sandra ORCID logo; Kirsch, Moritz ORCID logo; Ghamisi, Pedram (ORCID: 0000-0003-1203-741X); Tolosana Delgado, Raimon ORCID logo; Gloaguen, Richard (ORCID: 0000-0002-4383-473X); Heizmann, Michael ORCID logo
Publisher Rodare
Publication Year 2023
Rights Creative Commons Attribution 4.0 International; Open Access; https://creativecommons.org/licenses/by/4.0/legalcode; info:eu-repo/semantics/openAccess
OpenAccess true
Contact https://rodare.hzdr.de/support
Representation
Language English
Resource Type Dataset
Discipline Life Sciences; Natural Sciences; Engineering Sciences