resolution Satellite Data Candidate Miss Thitirat Wiyarat Special Project Studies Advisor Dr. Tiranee Achalakul

This paper discusses a design framework of the fusion of the multi-spectral image data for the analysis of satellite images. In recent years, as the availability of remote sensing imagery has increased, fusing images of different wavelength and different resolution have become a significant operation in remote sensing study. Image fusion process is designed to enhance the resolution of multi-spectral images by merging a relatively coarse-resolution image with a higher resolution of the same geographic area. The result of the fusion process is a single image which contains more information for human vision and machine perception. We have designed and implemented the fusion model for the satellite image analysis in this work. The model is based on a multi-scale decomposition (wavelet transform) and the maximum coefficient fusion rule. To evaluate the model, it was applied to the multi-spectral image taken from the Landsat ETM+ sensor. The image of Chiangmai, Thailand is used as our experimental scene due to a good mixture of foliated and urban areas contained in the scene. The resulting image quality was evaluated using the SNR and the RMSE as well as human interpretation