WTS: A Weakly towards Strongly Supervised Learning Framework for Remote Sensing Land Cover Classification Using Segmentation Models

WTS: A Weakly towards Strongly Supervised Learning Framework for Remote Sensing Land Cover Classification Using Segmentation Models

Land cover classification is one of the most fundamental tasks in the field of remote sensing. In recent years, fully supervised fully convolutional network (FCN)-based semantic segmentation models have achieved state-of-the-art performance in the semantic segmentation task. However, creating pixel-...

Descripción completa

Detalles Bibliográficos
Autores principales: Wei Zhang, Ping Tang, Thomas Corpetti, Lijun Zhao
Formato: Artículo
Lenguaje:English
Publicado: MDPI AG 2021-01-01
Colección:Remote Sensing
Materias:
land cover classification
convolutional neural network
segmentation model
weakly supervised
seeded region growing
fully connected Conditional Random Fields
Acceso en línea:https://www.mdpi.com/2072-4292/13/3/394

Internet

https://www.mdpi.com/2072-4292/13/3/394

Ejemplares similares