Orthogonalization methods for multispectral data of remote sensing 
of the Earth are considered. Four particular orthogonalization process are considered: 
the principal components method, Gram-Schmidt orthogonalization, QR-orthogonalization, 
singular value decomposition. New method of increasing of the spatial resolution 
of multispectral images is proposed. Method can simultaneously increase 
the spatial resolution of the image in all spectral bands. 
Increasing of the spatial resolution is based on the use of orthogonalization process 
and panchromatic band data. Quantitative evaluation of these methods is performed.
Keywords: multispectral image, orthogonalization, the principal components method,
 Gram-Schmidt orthogonalization, entropy, QR-orthogonalization, singular value decomposition.