The accuracy in classification of remote sensing (RS) images using deep learning architectures is affected by the lack of large sets of training samples. Although a significant effort is currently devoted to generate databases of annotated satellite images, these datasets may not be large enough to accurately model at global level different types of land-cover surfaces. To solve such a problem, this paper presents an unsupervised approach which aims to exploit the RS image that has to be classified and publicly available thematic products to generate a training database of weak samples representative of the considered study area. First, we harmonize the thematic map and the RS image. Then, samples having the highest probability to be correctly associated to their labels are extracted from the map by exploiting the information provided by the RS image to be classified. Finally, the weak labeled samples are used to train a convolutional neural network (CNN). Experimental results obtained training a CNN on Sentinel 2 images with weak labels extracted from the 2018 corine land cover (CLC) map demonstrate the effectiveness of the proposed method.
Automatic Extraction of Weak Labeled Samples From Existing Thematic Products For Training Convolutional Neural Networks / Paris, Claudia; Bruzzone, Lorenzo. - (2019), pp. 5722-5725. (Intervento presentato al convegno IGARSS 2019 tenutosi a Yokohama nel 28th July-2nd August 2019) [10.1109/IGARSS.2019.8900649].
Automatic Extraction of Weak Labeled Samples From Existing Thematic Products For Training Convolutional Neural Networks
Paris, Claudia;Bruzzone, Lorenzo
2019-01-01
Abstract
The accuracy in classification of remote sensing (RS) images using deep learning architectures is affected by the lack of large sets of training samples. Although a significant effort is currently devoted to generate databases of annotated satellite images, these datasets may not be large enough to accurately model at global level different types of land-cover surfaces. To solve such a problem, this paper presents an unsupervised approach which aims to exploit the RS image that has to be classified and publicly available thematic products to generate a training database of weak samples representative of the considered study area. First, we harmonize the thematic map and the RS image. Then, samples having the highest probability to be correctly associated to their labels are extracted from the map by exploiting the information provided by the RS image to be classified. Finally, the weak labeled samples are used to train a convolutional neural network (CNN). Experimental results obtained training a CNN on Sentinel 2 images with weak labels extracted from the 2018 corine land cover (CLC) map demonstrate the effectiveness of the proposed method.File | Dimensione | Formato | |
---|---|---|---|
paris.pdf
Solo gestori archivio
Tipologia:
Post-print referato (Refereed author’s manuscript)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
217.85 kB
Formato
Adobe PDF
|
217.85 kB | Adobe PDF | Visualizza/Apri |
08900649.pdf
Solo gestori archivio
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
242.6 kB
Formato
Adobe PDF
|
242.6 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione