Advanced search

Segmentation of images using a simplified watershed algorithm

Author
Organization
Abstract
The watershed algorithm is a technique used for image segmentation. The segmentation here signifies that the image is partitioned into connected regions. Each region contains pixels having more or less the same characteristics. A region may be characterized by a mean value of gray levels, variances of gray levels, etc. The standard watershed algorithm, based on the use of mathematical morphology was proposed in (1). The current paper describes a modification of this algorithm, which is easier to explain and implement, and gives satisfactory results in a typical case. The paper presents two applications of the watershed algorithm. One of them is related to the quality control, and in particular to the detection of defects of the manufactured ferrite cores. The watershed algorithm is then particularly suited to the extraction of the shape of large defects. The other presented application of the watershed is in medical imaging. As an example, the use of the watershed in the extraction of the gray matter in an NMR image of a human spinal cord is presented.

Citation

Please use this url to cite or link to this publication:

Chicago
Serneels, R, N Nieniewski, and Etienne Kerre. 2000. “Segmentation of Images Using a Simplified Watershed Algorithm.” In Intelligent Techniques and Soft Computing in Nuclear Science and Engineering, ed. D Ruan, HA Abderrahim, P D’Hondt, and Etienne Kerre, 231–238. Singapore, Singapore: World Scientific.
APA
Serneels, R., Nieniewski, N., & Kerre, E. (2000). Segmentation of images using a simplified watershed algorithm. In D Ruan, H. Abderrahim, P. D’Hondt, & E. Kerre (Eds.), Intelligent techniques and soft computing in nuclear science and engineering (pp. 231–238). Presented at the 4th International FLINS conference on Intelligent Techniques and Soft Computing in Nuclear Science and Engineering, Singapore, Singapore: World Scientific.
Vancouver
1.
Serneels R, Nieniewski N, Kerre E. Segmentation of images using a simplified watershed algorithm. In: Ruan D, Abderrahim H, D’Hondt P, Kerre E, editors. Intelligent techniques and soft computing in nuclear science and engineering. Singapore, Singapore: World Scientific; 2000. p. 231–8.
MLA
Serneels, R, N Nieniewski, and Etienne Kerre. “Segmentation of Images Using a Simplified Watershed Algorithm.” Intelligent Techniques and Soft Computing in Nuclear Science and Engineering. Ed. D Ruan et al. Singapore, Singapore: World Scientific, 2000. 231–238. Print.
@inproceedings{123670,
  abstract     = {The watershed algorithm is a technique used for image segmentation. The segmentation here signifies that the image is partitioned into connected regions. Each region contains pixels having more or less the same characteristics. A region may be characterized by a mean value of gray levels, variances of gray levels, etc. The standard watershed algorithm, based on the use of mathematical morphology was proposed in (1). The current paper describes a modification of this algorithm, which is easier to explain and implement, and gives satisfactory results in a typical case. The paper presents two applications of the watershed algorithm. One of them is related to the quality control, and in particular to the detection of defects of the manufactured ferrite cores. The watershed algorithm is then particularly suited to the extraction of the shape of large defects. The other presented application of the watershed is in medical imaging. As an example, the use of the watershed in the extraction of the gray matter in an NMR image of a human spinal cord is presented.},
  author       = {Serneels, R and Nieniewski, N and Kerre, Etienne},
  booktitle    = {Intelligent techniques and soft computing in nuclear science and engineering},
  editor       = {Ruan, D and Abderrahim, HA and D'Hondt, P and Kerre, Etienne},
  isbn         = {9789810243562},
  language     = {eng},
  location     = {Brugge, Belgium},
  pages        = {231--238},
  publisher    = {World Scientific},
  title        = {Segmentation of images using a simplified watershed algorithm},
  year         = {2000},
}

Web of Science
Times cited: