Fast, parallel watershed algorithm based on path tracing

Authors:
Michał Świercz;Marcin Iwanowski
Affiliations:
Institute of Control and Industrial Electronics, Warsaw University of Technology, Warsaw, Poland;Institute of Control and Industrial Electronics, Warsaw University of Technology, Warsaw, Poland
Venue:
ICCVG'10 Proceedings of the 2010 international conference on Computer vision and graphics: Part II
Year:
2010

Citing 8
Cited 1

Watersheds in Digital Spaces: An Efficient Algorithm Based on Immersion Simulations

IEEE Transactions on Pattern Analysis and Machine Intelligence
Parallel Image Component Labeling With Watershed Transformation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Parallel watershed transformation algorithms for image segmentation

Parallel Computing
The watershed transform: definitions, algorithms and parallelization strategies

Fundamenta Informaticae - Special issue on mathematical morphology
A fast watershed algorithm based on chain code and its application in image segmentation

Pattern Recognition Letters
An improved watershed algorithm based on efficient computation of shortest paths

Pattern Recognition
An efficient immersion-based watershed transform method and its prototype architecture

Journal of Systems Architecture: the EUROMICRO Journal
Comparison between immersion-based and toboggan-based watershed image segmentation

IEEE Transactions on Image Processing

Building programming abstractions for wireless sensor networks using watershed segmentation

NEW2AN'11/ruSMART'11 Proceedings of the 11th international conference and 4th international conference on Smart spaces and next generation wired/wireless networking

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper a fast, parallel watershed algorithm for segmentation of digital grey-scale images is presented. We show an original parallelisation technique based on the "shared nothing" principle and its application to a modified path-tracing watershed algorithm, which allows a vast majority of computations to be broken up into several independent tasks that can be run in parallel on different processing nodes. This approach eliminates the need of any complex synchronization between the threads and translates to a very high efficiency and speed of the algorithm. Sample results are discussed, with emphasis on their correctness and execution times.