Fast and robust curve skeletonization for real-world elongated objects

Authors: Tabb, Amy; MEDEIROS, HENRY - Marquette University

Submitted to: IEEE Winter Conference on Applications of Computer Vision
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/19/2018
Publication Date: 3/11/2018
Citation: Tabb, A., Medeiros, H. 2018. Fast and robust curve skeletonization for real-world elongated objects. IEE Winter Conference on Applications of Computer Vision. https://doi.org/10.1109/WACV.2018.00214.
DOI: https://doi.org/10.1109/WACV.2018.00214

Interpretive Summary: We considered the problem of computing the internal structure of three-dimensional models. We describe a method for computing the internal structure or curve skeleton that is fast and also able to deal with real-world conditions. This work will be used to compute the internal structure of three-dimensional models of trees for phenotyping.

Technical Abstract: We consider the problem of extracting curve skeletons of three-dimensional, elongated objects given a noisy surface, which has applications in agricultural contexts such as extracting the branching structure of plants. We describe an efficient and robust method based on breadth-first search that can determine curve skeletons in these contexts. The run time of our method is less than four seconds on large, challenging datasets, which makes it appropriate for situations where real-time decision making is needed. Our method compares favorably to classical thinning algorithms as well as to recent contributions to the field.