A SOM-based Chan-Vese Model for Unsupervised Image Segmentation
- Submitting institution
-
Birmingham City University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11Z_OP_D0029
- Type
- D - Journal article
- DOI
-
10.1007/s00500-015-1906-z
- Title of journal
- Soft Computing
- Article number
- -
- First page
- 2047
- Volume
- 21
- Issue
- 8
- ISSN
- 1432-7643
- Open access status
- Out of scope for open access requirements
- Month of publication
- -
- Year of publication
- 2015
- URL
-
https://link.springer.com/article/10.1007/s00500-015-1906-z
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Citation count
- 9
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The paper proposes a novel image segmentation method based on active contour model (ACM), combining the advantages of Self-Organizing Maps (SOM) within the level set framework of a state-of-the-art unsupervised global ACM, namely, the Chan-Vese (C-V ) model. Novel formulations for unsupervised active contour models, based on self-organizing neurons have been devised. Thorough experimentation on both real and synthetic images proved the effectiveness of the proposed model in segmenting regions of interest, in the presence of additive noise. Based on the highly accurate results reported, the model has inspired work in medical applications (e.g. DOI: 10.1016/j.asoc.2019.105547, and DOI: 10.1155/2019/7632308).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -