Deep Learning Architecture for Forest Detection in Satellite Data
Deep Learning algorithms have achieved great progress in different applications due to their training capabilities, parameter reduction and increased accuracy. Image processing is a particular area that has received recent attention promoted by the growing processing power and data availability. Rem...
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Objeto de conferencia |
| Lenguaje: | Inglés |
| Publicado: |
2019
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/90894 |
| Aporte de: |
| Sumario: | Deep Learning algorithms have achieved great progress in different applications due to their training capabilities, parameter reduction and increased accuracy. Image processing is a particular area that has received recent attention promoted by the growing processing power and data availability. Remote sensing devices provide image-like data that can be used to characterize Earth’s natural or artificial phenomena.
Particularly, forest detection is important in many applications like flooding simulations, analysis of forest health or detection of area desertification.
The existing techniques for forest detection based on satellite data lack accuracy or still require human expert intervention to correct recognition errors or parameter setup. In this work a Deep Learning architecture for forest detection is presented, that aims at increasing accuracy and reducing expert dependency. A data preprocessing procedure, analysis and dataset composition for robust automatic forest detection is described. The proposed approach was validated with real SRTM and Landsat-8 satellite data. |
|---|