ORTHOGONAL TRANSFORMATION OF SEGMENTED IMAGES FROM THE SATELLITE SENTINEL-2

The aim of this work is to develop a model for orthogonal transformation, called Tasselled Cap, to Sentinel-2 images. The essence of the Tasselled Cap Transformation (TCT) model is the unitary matrix for orthogonal transformation. In this paper is generated unitary matrix for ortogonalisation of the images from MSI (MultiSpectral Instrument) sensor Sentinel-2. Brightness, Greenness, and Wetness Tasselled Cap indices were derived for the MSI Sentinel-2 sensor. Eight images of sensor Sentinel-2 from July 2016 were used. Two images are from Northwestern Bulgaria, two are from the central part of Northern Bulgaria, two from Northeast Bulgaria, and the remaining two from the central and eastern part of Southern Bulgaria. Test areas are used to choose respective homogenous clusters for determining the base vector of Orthogonal Transformation, Brightness and the two other components, Greenness, and Wetness. Based on the obtained unitary matrix a model for automatic preprocessing of Sentinel-2 images was developed aimed at obtaining Brightness, Greenness, and Wetness Tasselled Cap indices. Part of the obtained results from applying the orthogonalisation model are presented for four test areas of MSI sensor Sentinel-2 images.