Survey on 3D Reconstruction Techniques: Large-Scale Urban City Reconstruction and Requirements

Andreas Christodoulides, Gary K. L. Tam, James Clarke, Richard Smith, Jon Horgan, Nicholas Micallef, Jeremy Morley, Nelly Villamizar, Sean P. Walton

Abstract:

3D representations of large-scale and urban scenes are
crucial across various industries, including autonomous driving,
urban planning, natural resource supervision and many more.
Large-scale industrial reconstructions are inherently complex and
multifaceted. Many existing surveys primarily focus on academic
progressions and often neglect the intricate and diverse needs
of industry. This survey aims to bridge this gap by providing
a comprehensive analysis of 3D reconstruction methods, with a
focus on industrial requirements such as scalability and integration
of human interaction. Our approach involves utilizing Affinity
Diagramming to systematically analyze qualitative data gathered
from industrial partners. This methodology enables us to gain deep
insights into how recent literature addresses these specific indus-
trial needs. The survey encompasses various aspects, including in-
put and reconstruction modalities, architectural models, datasets,
evaluation metrics, and the incorporation of prior knowledge. We
further discuss practical implications derived from our analysis,
highlighting key considerations for future advancements in 3D
reconstruction methods tailored for large-scale applications.