Comparison of fusion by morphological pyramid and by high pass filtering

Image fusion is the process of combining information from two or more images of a scene into single composite image that is more informative and more suitable for visual perception or computer processing. Given the same set of the input images, different images may be created depending on the specific application and what is considered relevant information. High resolution panchromatic (Pan) images provide better spatial quality compared to multispectral (MS) images. However MS images provide better spectral quality compared to Pan image. Objectif recognition and extraction of urban features require methods to integrate complementary data sets such as PAN and MS. Many pixel- based image fusion techniques have been developed to combine the spatial and spectral characteristics of Pan and MS images in synthesized MS images. In this paper, we introduce a method based on pyramid algorithms and morphological filtering to create synthesis images having both high spatial resolution and high spectral resolution. This method includes seven steps as well as geometric registration, decomposition by morphological pyramid, fusion and reconstruction by morphological pyramid. The fusion, in our case, consists to replace the approximation image of the high resolution image with the multispectral image (low resolution) using interpolation and add detail information to reconstruction apply interpolation and add details information in each levels. Then this method is compared into the HPF (High Pass Filtering) method. In the HPF, the high frequencies in Pan are extracted using high pass filters. The extracted high frequencies are then introduced into one band of the MS image by simple addition. Two approaches have been used to compare and evaluate the fusion products: the first one is a quantitative analysis, performed with the calculation of the mathematical indicators (the bias, the correlation coefficient and ERGAS (means relative global dimensional synthesis error)). The second one- is a qualitative analysis by visual experts.