Comparison of 3D Texture-based Image Descriptors in Fluorescence Microscopy
| Authors | |
|---|---|
| Year of publication | 2014 |
| Type | Article in Proceedings |
| Conference | 16th International Workshop on Combinatorial Image Analysis (IWCIA) 2014 |
| MU Faculty or unit | |
| Citation | |
| Web | http://link.springer.com/chapter/10.1007/978-3-319-07148-0_17 |
| Doi | http://dx.doi.org/10.1007/978-3-319-07148-0_17 |
| Field | Use of computers, robotics and its application |
| Keywords | 3D images; Texture descriptors; Fluorescence microscopy; Local Binary Patterns |
| Description | In recent years, research groups concentrate still more attention on 3D images, especially in the field of biomedical image processing. Adding another dimension enables to capture the entire object. On the other hand, handling 3D images requires also new algorithms, since not all of them can be modified for higher dimensions intuitively. In this article, we introduce a comparison of various implementations of 3D texture descriptors presented in the literature in recent years. We prepared an unified environment to test all of them under the same conditions. From the results of our tests we came to conclusion, that 3D variants of LBP in the combination with k-NN classifier are very strong approach with the classification accuracy more than 99% on selected group of 3D biomedical images. |
| Related projects: |