A Segmentation Technique for Flexible Pipes in Deep Underwater Environments

Saulo Pessoa, Vinicius Cesar, Bernardo Reis, Judith Kelner and Ismael Santos

Abstract

This paper presents a segmentation technique for flexible pipes in deep underwater environments using low-light monochrome cameras. The technique relies on an alternating pattern of black and white regions marked over the pipe and is divided into three stages: a pre-processing stage for image noise-reduction; a multi-level topological binarization for collecting pipe region candidates; and a backtracking search constrained by inherent pipe characteristics for segmenting its regions. The proposed technique has been tested using video sequences from a real offshore operation and succeeded in segmenting 95.29% of the frames, while local adaptive thresholding methods achieved, at best, a rate of 68.49%.

Session

Poster 2

Files

PDF iconExtended Abstract (PDF, 385K)
PDF iconPaper (PDF, 1650K)
ZIP iconSupplemental Materials (ZIP, 26M)

DOI

10.5244/C.29.135
https://dx.doi.org/10.5244/C.29.135

Citation

Saulo Pessoa, Vinicius Cesar, Bernardo Reis, Judith Kelner and Ismael Santos. A Segmentation Technique for Flexible Pipes in Deep Underwater Environments. In Xianghua Xie, Mark W. Jones, and Gary K. L. Tam, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 135.1-135.12. BMVA Press, September 2015.

Bibtex

@inproceedings{BMVC2015_135,
	title={A Segmentation Technique for Flexible Pipes in Deep Underwater Environments},
	author={Saulo Pessoa and Vinicius Cesar and Bernardo Reis and Judith Kelner and Ismael Santos},
	year={2015},
	month={September},
	pages={135.1-135.12},
	articleno={135},
	numpages={12},
	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
	publisher={BMVA Press},
	editor={Xianghua Xie, Mark W. Jones, and Gary K. L. Tam},
	doi={10.5244/C.29.135},
	isbn={1-901725-53-7},
	url={https://dx.doi.org/10.5244/C.29.135}
}