A Comprehensive Review of Advances in 3D Point Cloud Generation and Mesh Texturing
IEEE Access · 2026 · 14 · 98752-98777
DOI: 10.1109/access.2026.3707495 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Giovanna Bueno Marcondes und 5 weiteren Autor:innen. Die Arbeit nennt 65 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: VirtuAuto Project sponsored by FINEP - Partners: PUC Minas, Stellantis, and UTFPR-PG.
Abstract (Originalsprache)
Three-dimensional (3D) point cloud generation and mesh texturing have become fundamental components in computer vision, graphics, and image processing, enabling high-fidelity representations of real-world environments. These techniques form the backbone of applications such as autonomous navigation, virtual and augmented reality, cultural heritage preservation, medical imaging, and large-scale urban modeling. Despite rapid progress, the field remains fragmented, with various methodologies lacking a unified comparative framework that highlights trade-offs, limitations, and emerging opportunities. This paper presents a comprehensive survey supported by a structured PRISMA-inspired review methodology to ensure rigor and reproducibility. We classify and analyze traditional geometry-based approaches (e.g., structure-from-motion, multi-view stereo, and Delaunay triangulation) alongside recent neural and hybrid methods (e.g., convolutional networks, neural radiance fields, and Gaussian splatting). Strengths and limitations are examined with respect to accuracy, efficiency, scalability, and robustness across varied scenarios, including including urban-scale, outdoor, and real-time reconstruction scenarios, large-scale cities, and real-time applications. Furthermore, we provide a comparative synthesis of state-of-the-art tools, datasets, and benchmarks, discuss challenges such as sensor fusion, texture alignment, and computational complexity, and identify promising research avenues for next-generation reconstruction pipelines. By bridging the full spectrum from point cloud generation to mesh texturing, this review offers researchers and practitioners a comprehensive reference to guide methodological choices and foster innovation in 3D reconstruction.
Förderung
- VirtuAuto Project sponsored by FINEP - Partners: PUC Minas, Stellantis, and UTFPR-PG
Zitieren
APA
Marcondes, G. B., Hoffmann, J. E., Santos, M. M. D., Malik, A. W., Netto, M. L., & Justo, J. F. (2026). A Comprehensive Review of Advances in 3D Point Cloud Generation and Mesh Texturing. IEEE Access, 14, 98752-98777. https://doi.org/10.1109/access.2026.3707495
Vancouver
Marcondes GB, Hoffmann JE, Santos MMD, Malik AW, Netto ML, Justo JF. A Comprehensive Review of Advances in 3D Point Cloud Generation and Mesh Texturing. IEEE Access. 2026;14:98752-98777. doi:10.1109/access.2026.3707495
BibTeX
@article{Marcondes2026,
author = {Marcondes, Giovanna Bueno and Hoffmann, João Eduardo and Santos, Max Mauro Dias and Malik, Asad Waqar and Netto, Marcio Lobo and Justo, João Francisco},
title = {A Comprehensive Review of Advances in 3D Point Cloud Generation and Mesh Texturing},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {98752--98777},
doi = {10.1109/access.2026.3707495}
}
EUVIMED