A correlation comparison for remotely sensed data with different resolutions

More and more high resolution images are used in the process of land use dynamic monitoring and object recognizing. Different resolution images have different capabilities in land recognition and classification. The high resolution images play an important role in this area. This paper uses a model to analyze the spatial correlation among the images of different resolutions. The spatial correlations of different types are calculated and compared within this model. As the analysis illustrates, the spatial correlation decreases with the spatial extension for the city land, whereas it is the reverse case for the plantation under the same condition. Furthermore, the research study on the different resolution images for the same area shows that the images dimensions are varied with the spatial resolution and the spatial correlations are changed with different dimensions. The spatial correlation of large-dimension images is greater than that of small-dimension images. It means the spatial correlation of high resolution images is a little bit higher than that of low resolution images. This study reveals that the spatial correlation can be used in evaluating the information entropy, and is very helpful in object recognizing and classifying.