Long-Term Multi-Session 3D Reconstruction Under Substantial Appearance Change

TL;DR

This paper presents a novel joint SfM approach that uses combined handcrafted and learned features to enable coherent 3D reconstruction across long-term, appearance-changing datasets like coral reefs, outperforming existing methods.

Contribution

It introduces a joint SfM pipeline with cross-session correspondences and a scalable matching strategy for long-term environmental monitoring.

Findings

Successful reconstruction across years with significant change

Robust cross-session feature matching improves alignment

Scalable approach reduces computational cost

Abstract

Long-term environmental monitoring requires the ability to reconstruct and align 3D models across repeated site visits separated by months or years. However, existing Structure-from-Motion (SfM) pipelines implicitly assume near-simultaneous image capture and limited appearance change, and therefore fail when applied to long-term monitoring scenarios such as coral reef surveys, where substantial visual and structural change is common. In this paper, we show that the primary limitation of current approaches lies in their reliance on post-hoc alignment of independently reconstructed sessions, which is insufficient under large temporal appearance change. We address this limitation by enforcing cross-session correspondences directly within a joint SfM reconstruction. Our approach combines complementary handcrafted and learned visual features to robustly establish correspondences across large temporal gaps, enabling the reconstruction of a single coherent 3D model from imagery captured years apart, where standard independent and joint SfM pipelines break down. We evaluate our method on long-term coral reef datasets exhibiting significant real-world change, and demonstrate consistent joint reconstruction across sessions in cases where existing methods fail to produce coherent reconstructions. To ensure scalability to large datasets, we further restrict expensive learned feature matching to a small set of likely cross-session image pairs identified via visual place recognition, which reduces computational cost and improves alignment robustness.