Authors:
Kawtar El Karfi
1
;
Sanaa El Fkihi
1
;
Loubna El Mansouri
2
and
Othmane Naggar
3
Affiliations:
1
ENSIAS, Morocco
;
2
IAV
;
3
Mascir, Morocco
Keyword(s):
Remote sensing data, Hyperspectral image HSI, Classification,Crop type mapping ,Machine learning,Deep learning
Abstract:
Hyperspectral imagery (HSI) is widely considered to be one of the most used technologies in different remote sensing applications, such as crop mapping, which provides an essential baseline for understanding and monitoring the Earth. Hyperspectral remote sensing, with its multiple narrow and continuous wavebands, allow significant improvements in the understanding of physiological processes of crops and the changes in their phenology, which are indistinct in multi-spectral remote sensing. A generous number of features can be derived from the hyperspectral data, although the classification of crops using high-dimensional and high-resolution data is a challenging task. The main objective of this paper is to list various techniques of machine learning mostly applied for hyperspectral data classification, besides the different hyperspectral open datasets mainly used in various researches.