Zailskas, Felix (2022) An Evaluation of three Dissimilarity Measures for Alpha Trees in Colored Images. Bachelor's Thesis, Computing Science.
|
Text
bCS_2022_ZailskasF.pdf Download (14MB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
toestemming.pdf Restricted to Registered users only Download (124kB) |
Abstract
The α-tree algorithm is used for image segmentation by comparing and grouping adjacent pixels in an image based on a dissimilarity measure. This paper expands on previous research by investigating the Manhattan distance, the Euclidean distance, and the cosine dissimilarity as dissimilarity measures for images in the RGB color space. The dissimilarity measures are first tested on synthetic test images and then used to segment four satellite images. For the synthetic images a quantitative comparison is used to predict results on the satellite data. The resulting segmentation images of the satellite data are compared by eye to ground truth segmentation images with different detail levels at multiple α levels. Appropriate α levels were in the range of [60, 80] for the Manhattan distance, [70, 110] for the Euclidean distance, and [3, 15] for the cosine dissimilarity.
| Item Type: | Thesis (Bachelor's Thesis) |
|---|---|
| Supervisor name: | Wilkinson, M.H.F. and Bunte, K. |
| Degree programme: | Computing Science |
| Thesis type: | Bachelor's Thesis |
| Language: | English |
| Date Deposited: | 26 Aug 2022 14:16 |
| Last Modified: | 26 Aug 2022 14:16 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/28533 |
Actions (login required)
 |
View Item |