A novel oil spill feature selection and classification technique is presented, based on a forest of decision trees. The parameters of the two-class classification problem of oil spills and ¿look-alikes¿ have been reduced, selecting the most important features while increasing the true recognition rate. A series of tests have been conducted and results showed increased performance in oil spill recognition compared with previous studies. It is shown that from the initial 25 parameters used in these tests, only 6 are sufficient for robust recognition, which is also enhances the speed of processing.
PSYLLOS Apostolos;
TOPOUZELIS Konstantinos;
2014-08-26
ELSEVIER SCIENCE BV
JRC60404
0924-2716,
http://www.sciencedirect.com/science/article/pii/S0924271612000329,
https://publications.jrc.ec.europa.eu/repository/handle/JRC60404,
10.1016/j.isprsjprs.2012.01.005,
| Name | Country | City | Type |
|---|
This document is only visible at the Commission level.
You are not authorized to publish or distribute it outside the European Commission.
This is a public document. You can share this publication.
Datasets
| ID | Title | Public URL |
|---|
Dataset collections
| ID | Acronym | Title | Public URL |
|---|
Scripts / source codes
| Description | Public URL |
|---|
Additional supporting files
| File name | Description | File type |
|---|