| Title |
Skaitmeninių fotografijų palyginimo metodų tyrimas / |
| Translation of Title |
Investigation of methods for digital photography comparison. |
| Authors |
Matulionis, Marius |
| Full Text |
|
| Pages |
97 |
| Keywords [eng] |
computer vision ; algorithms ; photo comparison |
| Abstract [eng] |
In this paper we compare computer vision algorithms that can be used to compare digital photographs. We can separate photo comparison algorithms into two separate categories: pixel based algorithms and feature based algorithms. In this paper we focus on feature based algorithms because they are much more versatile and can be used to compare photos that have different resolotions and variuos distrortions which can not be compared with regular pixel based algorithms. But because feture based algorithms are so versatile, their abilities and weaknesses aren't fully reseached yet. This paper consist of analysis and experimentation parts. In the analysis part we look at photo comparison software and algorithms that they use to compare digital photographs. In the experimentation part we plan our experiments, write the conditions for the experiments and then test the selected algorithms. After completing the experiments we got data on how fast the algorithms work, how resistant they are to variuos distortions and photograph modifications. The algorithms with had the best run time and showed the most consistent accuracy was ORB algorithm. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2017 |
