FFT regression and cross-noise reduction for comparing images in remote sensing

摘要: In many remote sensing studies it is desired to quantify the functional relationship between images of a given target that were acquired by different sensors. Such comparisons are problematic because when pixel values one image plotted versus other, 'cross-noise' quite high. Typically, correlation coefficient low, even compared look alike. Nevertheless, we can try two suitable regression model function Y = f ( X ), while choosing them as 'the reference' and using other 'predictor' . The underlying assumption classical absolutely correct erroneous. Thus, objective fit parameters which minimize 'residuals' “ - ). When comparing in this not valid itself error prone. alternative FFT method presented herein comprises two...