Hi, welcome everyone! Good to see Geo-informatic community under one hub. Finally we got the platform where the activities/ achievements of various image processing projects can be shared. Moreover, I think this portal may cater the solutions to the problems faced during the implementation of Image Processing/ GIS activities. It’s a world known fact that satellite data plays an important and crucial role in the field of GIS. because its only Remote Sensing, which can provide the accurate output to be served as an input to GIS.
NIC is one of the premier organization in India, who have worked on a variety of IRS data products starting from WiFS (Resolution 188m) to cartosat1 (Resolution 2.5m) for various department i.e. water resources, power sector, urban planning, health, school, disaster and many more.
Biggest challenge in the field of Image processing is the accurate Geometric Correction of the satellite data received from NRSC, Hyderabad. Till the resolution of 5.8m(IRS-1D), we have achieved the target of Geometric Correction of the scenes across India with the help of SOI toposheets.
With the advent of IRS high resolution satellite data (2.5m), a bit of challenges were observed for its geometrical correction. First challenge was to find out the source of GCPs. For such high resolution satellite data, a source of approx. 10k scale is required. But there is no authenticated source of GCPs at this scale in the country. Highest available scale of SOI toposheet is 25K which is not appropriate source. A whisper in the corridor was that SOI shall complete the publishing of the toposheets at 10K scale for entire country by 2017, but as of now nothing seems available for the community.
Secondly the terrain corrected data received from NRSC, could not be used as it is, because it’s not precisely corrected and is having a quite mismatch with the other global services i.e. Google/Bing etc. NRSC also raised question on the georeferecing aspect of Google/ Bing which created a confusion on the authenticity and accuracy of the available input GCPs source.
An experiment was performed with the mobile GPS and found that the navigation exercise better matched with Google/Bing as compared to IRS terrain corrected data. Keeping this in view, the images were geo-referenced with these global services. In this way, we are in the process of developing the GCPs library which can be shared in GeoIT community.