Analysis of the identification of the digital images of the earth's Surface

Abstract

The quality of the identification of the digital images of Earth’s surface by using covariation functions and Wavelet (small wave) theory was investigated. Autocovariation function values of two digital images or single images are calculated using random functions, which are made of digital images pixels with vectors form. Pixel vector values are made scattering digital images pixels arrays to one row. For researches was used HJ-DEM IKONOS digital images, when cartographic projection is not orthographic and when scales of digital image points are different. Changing digital image scale single pixels colors waves frequencies often stay invariable, it means that scale changes not impact on Const and on covariation functions counting procedures. RGB color spectrum was used for images identification. Was analyzed RGB spectrum components and colors tensor influence to digital images covariation functions values. Identity of digital images is evaluated correlations coefficients values rotation at values range using software.