Sat-Mesh: Learning Neural Implicit Surfaces for Multi-View Satellite Reconstruction

Automatic reconstruction of surfaces from satellite imagery is a hot topic in computer vision and photogrammetry. State-of-the-art reconstruction methods typically produce 2.5D elevation data. In contrast, we propose a one-stage method directly generating a 3D mesh model from multi-view satellite im...

Full description

Saved in:
Bibliographic Details
Published in:Remote sensing (Basel, Switzerland) Vol. 15; no. 17; p. 4297
Main Authors: Qu, Yingjie, Deng, Fei
Format: Journal Article
Language:English
Published: Basel MDPI AG 01.08.2023
Subjects:
Accuracy
Artificial intelligence
Cameras
Comparative analysis
Computer architecture
Computer vision
Continuity (mathematics)
Efficiency
Finite element method
Geometry
Image enhancement
Image reconstruction
latent appearance
Learning
Machine learning
Mesh generation
Methods
neural implicit surfaces
Neural networks
neural radiance fields
normalized cross-correlation
Orientation behavior
Photogrammetry
Production methods
Remote sensing
satellite 3D reconstruction
Satellite imagery
Satellite imaging
satellites
Seasonal variations
Surface properties
surface quality
Technology application
texture
Three dimensional models
China
ISSN:2072-4292, 2072-4292
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Automatic reconstruction of surfaces from satellite imagery is a hot topic in computer vision and photogrammetry. State-of-the-art reconstruction methods typically produce 2.5D elevation data. In contrast, we propose a one-stage method directly generating a 3D mesh model from multi-view satellite imagery. We introduce a novel Sat-Mesh approach for satellite implicit surface reconstruction: We represent the scene as a continuous signed distance function (SDF) and leverage a volume rendering framework to learn the SDF values. To address the challenges posed by lighting variations and inconsistent appearances in satellite imagery, we incorporate a latent vector in the network architecture to encode image appearances. Furthermore, we introduce a multi-view stereo constraint to enhance surface quality. This constraint minimizes the similarity between image patches to optimize the position and orientation of the SDF surface. Experimental results demonstrate that our method achieves superior visual quality and quantitative accuracy in generating mesh models. Moreover, our approach can learn seasonal variations in satellite imagery, resulting in texture mesh models with different and consistent seasonal appearances.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2072-4292
2072-4292
DOI:10.3390/rs15174297