Geostatistical digital image merging

The merging of multisensor image data is becoming a widely used procedure because of the complementary nature of various data sets. Ideally, the method used to merge data sets with high-spatial and high-spectral resolution that should not distort the spectral characteristics of the high-spectral resolution data. The paper presents a geostatistical image merging method. The approach takes into consideration important aspects like, for example, support of information and makes a real image integration using a sequential gaussian conditional cosimulation-based procedure. All the parameters of the integration (basically, the weights corresponding to the images to merge) are extracted from the images themselves, providing additional information of them like variability structures that can be used in other digital image processes like filtering, texture determination and classification.