Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 4.76 MB
Add to list
  1. Search results
  2. Performing particle image segmentatio...

Performing particle image segmentation on an extremely small dataset

Marianna Chatzakou (UGent) , Junqing Huang (UGent) , Bogdan Parakhonskiy (UGent) , Michael Ruzhansky (UGent) , Andre Skirtach (UGent) , Junnan Song (UGent) and Xuechao Wang (UGent)
(2024) Extended Abstracts 2021/2022 : Ghent analysis and PDE seminar. In Trends in Mathematics 2. p.295-304
Author
Organization
Project
Abstract
Image segmentation is one of the typical computer vision tasks that has received great success with the recent advance of deep-learning methods. However, it is still a challenging problem, particularly when encountering limited data. In this paper, we present a new strategy for particle image segmentation, which relies on extensive data augmentation methods to reuse the available annotated samples for more effective performance. The procedure consists of the K-nearest neighbour (KNN) matting to fine-tune manually annotated boundaries and the use of data augmentations to expand available annotated data. After that, we employ the U-net architecture to train the model based on a small dataset consisting of twenty images. The results showed that the proposed strategy could effectively extract particle boundary features, thereby obtaining accurate segmentation results based on a very limited source of images.

Downloads

  • Download
    (...).pdf
    • full text (Accepted manuscript)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 4.76 MB

Citation

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

MLA
Chatzakou, Marianna, et al. “Performing Particle Image Segmentation on an Extremely Small Dataset.” Extended Abstracts 2021/2022 : Ghent Analysis and PDE Seminar, edited by Michael Ruzhansky and Karel Van Bockstal, vol. 2, Birkhäuser, 2024, pp. 295–304, doi:10.1007/978-3-031-42539-4_33.
APA
Chatzakou, M., Huang, J., Parakhonskiy, B., Ruzhansky, M., Skirtach, A., Song, J., & Wang, X. (2024). Performing particle image segmentation on an extremely small dataset. In M. Ruzhansky & K. Van Bockstal (Eds.), Extended Abstracts 2021/2022 : Ghent analysis and PDE seminar (Vol. 2, pp. 295–304). https://doi.org/10.1007/978-3-031-42539-4_33
Chicago author-date
Chatzakou, Marianna, Junqing Huang, Bogdan Parakhonskiy, Michael Ruzhansky, Andre Skirtach, Junnan Song, and Xuechao Wang. 2024. “Performing Particle Image Segmentation on an Extremely Small Dataset.” In Extended Abstracts 2021/2022 : Ghent Analysis and PDE Seminar, edited by Michael Ruzhansky and Karel Van Bockstal, 2:295–304. Cham: Birkhäuser. https://doi.org/10.1007/978-3-031-42539-4_33.
Chicago author-date (all authors)
Chatzakou, Marianna, Junqing Huang, Bogdan Parakhonskiy, Michael Ruzhansky, Andre Skirtach, Junnan Song, and Xuechao Wang. 2024. “Performing Particle Image Segmentation on an Extremely Small Dataset.” In Extended Abstracts 2021/2022 : Ghent Analysis and PDE Seminar, ed by. Michael Ruzhansky and Karel Van Bockstal, 2:295–304. Cham: Birkhäuser. doi:10.1007/978-3-031-42539-4_33.
Vancouver
1.
Chatzakou M, Huang J, Parakhonskiy B, Ruzhansky M, Skirtach A, Song J, et al. Performing particle image segmentation on an extremely small dataset. In: Ruzhansky M, Van Bockstal K, editors. Extended Abstracts 2021/2022 : Ghent analysis and PDE seminar. Cham: Birkhäuser; 2024. p. 295–304.
IEEE
[1]
M. Chatzakou et al., “Performing particle image segmentation on an extremely small dataset,” in Extended Abstracts 2021/2022 : Ghent analysis and PDE seminar, Ghent, Belgium, 2024, vol. 2, pp. 295–304.
@inproceedings{01J5X1M604QV05HWGSYHXKMS3K,
  abstract     = {{Image segmentation is one of the typical computer vision tasks that has received great success with the recent advance of deep-learning methods. However, it is still a challenging problem, particularly when encountering limited data. In this paper, we present a new strategy for particle image segmentation, which relies on extensive data augmentation methods to reuse the available annotated samples for more effective performance. The procedure consists of the K-nearest neighbour (KNN) matting to fine-tune manually annotated boundaries and the use of data augmentations to expand available annotated data. After that, we employ the U-net architecture to train the model based on a small dataset consisting of twenty images. The results showed that the proposed strategy could effectively extract particle boundary features, thereby obtaining accurate segmentation results based on a very limited source of images.}},
  author       = {{Chatzakou, Marianna and Huang, Junqing and Parakhonskiy, Bogdan and Ruzhansky, Michael and Skirtach, Andre and Song, Junnan and Wang, Xuechao}},
  booktitle    = {{Extended Abstracts 2021/2022 : Ghent analysis and PDE seminar}},
  editor       = {{Ruzhansky, Michael and Van Bockstal, Karel}},
  isbn         = {{9783031425387}},
  issn         = {{2297-0215}},
  language     = {{eng}},
  location     = {{Ghent, Belgium}},
  pages        = {{295--304}},
  publisher    = {{Birkhäuser}},
  title        = {{Performing particle image segmentation on an extremely small dataset}},
  url          = {{http://doi.org/10.1007/978-3-031-42539-4_33}},
  volume       = {{2}},
  year         = {{2024}},
}
Altmetric
View in Altmetric