Structure-and-Motion Pipeline on a Hierarchical Cluster Tree

with R.Toldo, R. Gherardi, M. Farenzena


This work introduces a novel hierarchical scheme for computing Structure and Motion. The images are organized into a tree with agglomerative clustering, using a measure of overlap as the distance. The reconstruction then follows this tree from the leaves to the root. As a result, the problems is broken into smaller instances, which are then separately solved and combined. Compared to the standard sequential approach, this pipeline boost computational efficiency by one order of magnitude, it is independent from the initial pair of views, and copes better with drift problems.

Samantha sprouted the development of a commercial software, called 3DF Zephyr. See this blog for the latest developements.




Images Reconstruction Video

54 JPG images / 4.7 Mb file

Piazza Dante

39 JPG images / 5.3 Mb file

Piazza Erbe

259 JPG images / 24 Mb file

Piazza Bra

380 PNG images / 821 Mb file


309 PNG images / Images not available

Piazza Bra [NEW]

331 JPG images / 358 Mb file

San Giacomo

269 JPG images / 522 Mb file


92 JPG images / 432 Mb file
This dataset is courtesy of L. Barazzetti (Gicarus Lab).

  • not available

Reference Papers

  1. Toldo, R.; Gherardi, R.; Farenzena, M. and Fusiello, A. Hierarchical structure-and-motion recovery from uncalibrated images. In Computer Vision and Image Understanding 140: 127-143, 2015. (PDF)
  2. Riccardo Gherardi, Michela Farenzena, and Andrea Fusiello. Improving the efficiency of hierarchical structure-and-motion. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2010), pages 1594-1600, San Francisco, CA, 13-18 June 2010. (PDF)
  3. A. M.Farenzena, A.Fusiello, R. Gherardi. Structure-and-Motion Pipeline on a Hierarchical Cluster Tree. Proceedings of the IEEE International Workshop on 3-D Digital Imaging and Modeling, Kyoto, October 2009. (PDF)