Soccio, Annapia (2019) Feature extraction and segmentation of hyperspectral images. [Magistrali biennali]
Full text disponibile come:
This work proposes an approach for hyperspectral images segmentation without direct application of common clustering methods on the hyperspectral data. The proposed approach reduces the spectral dimension of the image, through principal component analysis, and its spatial dimension, through wavelet transform, in order to apply the clustering algorithm on a lower resolution version of the data and then train a classifier to label the high resolution image.
Solo per lo Staff dell Archivio: Modifica questo record