SUMMARY

Land cover map mapping of various types use conventional remote sensing methods as a very important part of the methodology. Use of optical remote sensing data is limited due to the high cloud cover in tropical area such as Viet Nam, accordingly, the microwave remote sensing may be considered as a solution. This study aims to develop a method to create a land cover map using microwave remote sensing in combination with optical data. The TerraSAR-X and Spot 5 image of Binh Thuan area inVietnam were chosen to be processed for this study. Some filters such as Lee, Forst, Gama and local Sigma has been applied to the radar image in order to reduce the noise. The radar and optical data were geo-rectified to VN-2000 coordinate system. Then, data fusion techniques such as NDVI, PCA and Color Composite (HIS and Brovey) has been applied to the radar and Spot 5 images. A supervised classification was applied to fusioned images to identify different types of land cover such as water, forest, residential area, etc. The fusioned images were also unsupervised classified. Finally, the classified result of data fusion was compared to the SPOT5 classification and manualy edited to establish land cover map. During the processing and mapping steps, the role of TerraSAR-X data in land cover mapping was estimated