Robust Frequency-Based Structure Extraction

04/19/2020 ∙ by Tomasz Piotr Kucner, et al. ∙ 0

We propose a method for measuring how well each point in an indoor 2D robot map agrees with the underlying structure that governs the construction of the environment. This structure scoring has applications for, e. g., easier robot deployment and Cleaning of maps. In particular, we demonstrate its effectiveness for removing clutter and artifacts from real-world maps, which in turn is an enabler for other map processing components, e. g., room segmentation. Starting from the Fourier transform, we detect peaks in the unfolded frequency spectrum that correspond to a set of dominant directions. This allows us to reconstruct a nominal reference map and score the input map through its correspondence with this reference, without requiring access to a ground-truth map.



There are no comments yet.


page 1

page 4

page 5

page 6

page 7

page 8

Code Repositories


(ROSE) RObust frequency-based Structure Extraction

view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.