Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2509.02342 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909766563921920 |
|---|---|
| author | Slepova, Ksenia Oiye, Ivan Etoku van Gijzen, Martin B. |
| author_facet | Slepova, Ksenia Oiye, Ivan Etoku van Gijzen, Martin B. |
| contents | Image denoising and image segmentation play essential roles in image processing. Partial differential equations (PDE)-based methods have proven to show reliable results when incorporated in both denoising and segmentation of images. In our work, we discuss a multi-stage PDE-based image processing approach. It relies upon the nonlinear diffusion for noise removal and clustering and region growing for segmentation. In the first stage of the approach, the raw image is computed from noisy measurement data. The second stage aims to filter out the noise using anisotropic diffusion. We couple these stages into one optimisation problem which allows us to incorporate a diffusion coefficient based on a presegmented image. The third stage performs the final segmentation of the image. We demonstrate our approach on both images for which the ground truth is known and on MR measurements made by an experimental, inexpensive scanner. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_02342 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multi-stage PDE-based image processing techniques for noisy MRI scans Slepova, Ksenia Oiye, Ivan Etoku van Gijzen, Martin B. Numerical Analysis Image denoising and image segmentation play essential roles in image processing. Partial differential equations (PDE)-based methods have proven to show reliable results when incorporated in both denoising and segmentation of images. In our work, we discuss a multi-stage PDE-based image processing approach. It relies upon the nonlinear diffusion for noise removal and clustering and region growing for segmentation. In the first stage of the approach, the raw image is computed from noisy measurement data. The second stage aims to filter out the noise using anisotropic diffusion. We couple these stages into one optimisation problem which allows us to incorporate a diffusion coefficient based on a presegmented image. The third stage performs the final segmentation of the image. We demonstrate our approach on both images for which the ground truth is known and on MR measurements made by an experimental, inexpensive scanner. |
| title | Multi-stage PDE-based image processing techniques for noisy MRI scans |
| topic | Numerical Analysis |
| url | https://arxiv.org/abs/2509.02342 |