Abstract

Currently, the use ofmulti-sensor and multi-temporal satellite imagery has been widely used in manyapplications where long-term and dense time series earth observation datasetare required to increase the performance of updating and detecting the changesfrom landscape. However, in order to accurately extract the information fromsatellite imagery at different time of acquisition, the prerequisite is tostandardize the spectral reflectance value to minimize the factors of changingthe reflection value on the satellite image in time domain. There are two commonapproaches of normalizing the spectral reflectance values: (1) absoluteapproach and (2) relative approach. These methods applied in standard remotesensing images to calibrate satellite images taken from different times anddifferent sensors. The paper presents the using of IR-MAD method, one of therelatively standard spectral normalization approaches to standardize Landsat-8and Sentine l-2 imagery.