AdaSfM: From Coarse Global to Fine Incremental Adaptive Structure from Motion

1National University of Singapore 2Segway-Ninebot 3Willand *Data would not be publicly available due to commercial restrictions Acknowledgement:
Appreciate Zihao (he was an intern during this work) finishing the match refinement module and Song Shu for the exhaustive evaluation. For commercial cooperation, contact Jianming Li and Tianning Yu

AdaSfM takes images and measurements from low-cost sensors as inputs. The view graph is refined by the result of global SfM. The absolute poses from the global SfM are used as priors in the subsequent local SfM process. The final reconstruction result is merged into the global SfM reference frame.


Despite the impressive results achieved by many existing Structure from Motion (SfM) approaches, there is still a need to improve the robustness, accuracy, and efficiency for large-scale scenes with many outlier matches and sparse view graphs. In this paper, we propose AdaSfM: a coarse-to-fine adaptive SfM approach that is scalable to large-scale and challenging datasets. Our approach first does a coarse global SfM which improves the reliability of the view graph by leveraging measurements from low-cost sensors such as Inertial Measurement Units (IMUs) and wheel encoders. Subsequently, the view graph is divided into sub-scenes that are refined in parallel by a fine local incremental SfM regularized by the result from the coarse global SfM to improve the camera registration accuracy and alleviate scene drifts. Finally, our approach uses a threshold-adaptive strategy to align all local reconstructions to the coordinate frame of global SfM. Extensive experiments on large-scale benchmark datasets show that our approach achieves state-of-the-art accuracy and efficiency.


Challenges in Visual Mapping

Top: Dynamic Objects
Bottom: Specular, Repetitive Textures
Top: Features Distributed Not Evenly
Bottom: Texture-less

Feature Matches After Refinement

Evaluation on 4Seasons Dataset

Effect of Augmented View Graph

Effect of Adaptive Flood Fill Image Partition

Effect of Introducing Global SfM

Evaluation on Self-Collected Outdoor Dataset

Comparison with COLMAP

More Visual Results


  author    = {Yu Chen, Zihao Yu, Shu Song, Tianning Yu, Jianming Li, Gim Hee Lee},
  title     = {AdaSfM: From Coarse Global to Fine Incremental Adaptive Structure from Motion},
  year      = {2022},