Lammers, Jeroen (2022) Finding an optimal dissimilarity measure for hierarchical segmentation of satellite images using alpha-trees. Master's Internship Report, Computing Science.
|
Text
mCS_2022_LammersJ.pdf Download (12MB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
Toestemming.pdf Restricted to Registered users only Download (130kB) |
Abstract
In this paper we will introduce a new method of computing the dissimilarity between pixels based on a combination of the forward difference, backward difference and central difference. We will compare this method against the Lp-norm dissimilarity measure which is augmented using edge detector and ridge detector signals. We will also introduce an approximation of the area score which will be used to determine the quality of the results. Based on the resulting training scores we can conclude that the newly proposed method preforms slightly worse compared to the other methods. All methods provide robust solution when considering the scores on similar input data. However, due to problems which most likely stem from the area score approximation we obtain over-merged results from our training process which raises questions about the usability of the results.
| Item Type: | Thesis (Master's Internship Report) |
|---|---|
| Supervisor name: | Wilkinson, M.H.F. and Bunte, K. |
| Degree programme: | Computing Science |
| Thesis type: | Master's Internship Report |
| Language: | English |
| Date Deposited: | 20 Dec 2022 14:09 |
| Last Modified: | 20 Dec 2022 14:09 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/29064 |
Actions (login required)
 |
View Item |