University of Twente Student Theses
Creating segmentation masks for cell segmentation using interactive segmentation model (SAM)
Girish Nair, Meenakshi (2024) Creating segmentation masks for cell segmentation using interactive segmentation model (SAM).
|
PDF
7MB |
| Abstract: | Cell segmentation is a crucial process in the biomedical field as the size, shape or number of cells can provide a plethora of information for medical diagnosis of many diseases\cite{1}. However, cell segmentation can be a difficult task to tackle owing to the irregular shapes and overlapping cells that lead to poor boundary distinction. In addition to the difficulty in the segmentation of cells, the availability of ground truth to train the model is a bottleneck due to the the resource intensive process of creating labeled segmentation masks. This paper implements an automatic cell segmentation model using Mask R-CNN trained on microscopic cell images dataset. The Segment Anything Model (SAM) will be used for interactive segmentation of images to produce the ground truth to train the model. The primary objective of this study is to analyse how SAM prompts can be used interactively in order to produce accurate segmentation masks to use as ground truth to train segmentation models like Mask R-CNN. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/100779 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page