Satellite image fusion with multiscale wavelet analysis for marine applications: preserving spatial information and minimizing artifacts (PSIMA)

摘要

Image fusion has many diverse purposes. There is no universal, quantitative performance measure to estimate image fusion quality. An essential principle is that the original application-specific information is preserved while artifacts are minimized in the final product. As an application of wavelet analysis to the fusion of images with significantly different spectral properties and pixel spacings, we propose a method that is based on multiscale wavelet analysis, thus preserving spatial information and minimizing artifacts (PSIMA). The information content of the two images is emphasized. With the PSIMA method, the images are fused while minimizing the degree of resampling. Therefore, the finest spatial information in the input images can be preserved and artifacts minimized in the fused product, independent of the fusion method. We demonstrate the PSIMA method using RADARSAT-1 scanning synthetic aperture radar (ScanSAR) and National Oceanic and Atmospheric Administration advanced very high resolution radiometer (NOAA AVHRR) images. This marine data set is a precursor to contemporaneous SAR and electro-optical data that will be available from the advanced synthetic aperture radar (ASAR) and medium-resolution imaging spectrometer (MERIS) sensors on the environmental satellite (ENVISAT). The results show that the PSIMA method is superior to conventional image fusion methods in terms of spatial information preservation and artifact rejection.