Topographic data bases in the product generation of IRS_P6 Liss-4 imagery

A continuous and autonomous co-registration and geo-location of image data from multiple sensors with widely varying view angles and resolution is one of the unique challenges of IRS-P6 data processing. The Liss-4 camera is capable of imaging in either multi-spectral or panchromatic modes with steerability up to +/− 27°. As the satellite sweeps longer strips with oblique viewing configuration, data product generation requires an elevation database to accommodate the errors due to terrain topography. SRTM, GTOPO30, DCW etc. are becoming hidden information to product generation schemes. In this paper, we analyze and justify the requirement of inclusion of a global DEM in the product generation system for Liss-4 data. In the first part of the study, after a brief introduction of sensor model, comparison of elevation from SRTM and GTOPO30 is made which leads towards a conclusion that GTOPO30 is adequate for improving the accuracy. The second part deals with the impact of GTOPO30 on geometric accuracy. Thirdly we analyze the terrain effects on band-to-band registration of Liss-4 multispectral data. A perfect band-to-band registration cannot be achieved without inclusion of digital elevation model. Many Liss_4 data sets with different tilts that cover terrain patches with flat, rolling and highly undulating nature are studied and the results reported.