Analysis of Uncertainties in Visualizing Remote Sensing and Spatial Data

The maintenance of satellite and spatial data quality is an important task and escalating interest amongst geospatial community. The visualization of perfect earth surface features by relative response of spectral signatures in the remotely sensed satellite data is unique domain. The present work deals to identify and visualize the certainty and uncertainties that are occurred during the interpretation of magnetite quartzite iron ore formations of Godumalai region of granulite region of Pre-Cambrian age of part of Salem district, Tamil Nadu state, India. The Godumalai is made up of mixed type of rocks such as gneisses, amphibolites, quartzite, chlorite-schists, feldspathic schists and granites. The iron ore formation forms in the crest of the hill with an eastwest trend. The applicable image processing techniques and associated put forth of field knowledge has been utilized in interpretation of IRS P6 LISS IV MX Satellite data to discriminate the iron formations of the study area. The visual interpretation of FCC of digital data shows that the formation have the raw pixel values of iron ore bands in the blue ranges 95 to 120, green ranges 85 to 125 and red ranges 55 to 75 and provided better delineation for iron formations. During this study, the image processing of Minimum Noise Fraction transform (MNF), band rationing, principal component analysis and contrast enhancements are carried out under the environment for visualizing images using ENVI 4.2 and ERDAS Imagine 8.7. The interpretation shows that the iron ore band exhibits in dark color, linear shape, fine texture and sharp contact with associated rock types. The dark bands arise suspicious due to the impact of hill shadows, hill ridges, sun azimuth and others. To rectify and reduce misinterpretation all above image processing techniques are undertaken along with the detailed filed-check in the selected test sites using GPS 178 and delineated the in situ iron formations. The observations show that the weathered formations can also be traced on the imagery along with massive formations since the absorption’s themselves are different. The availability spectral absorption of formation is inferred highly in near infrared region (0.77-0.86 μm). Similarly, the information content of 3D view is infinitely greater than the 2D perspective view, since it is viewing imagery beyond GIS requires to environmentalist, geologist and others. In which, the real model quality depends fistful of issues such as reference, range, reflection of the surface, reflectivity, resolution of the scanner and the reconstruction of the software. Under these circumstances, the visualization confirms few uncertainties in mapping of iron ore deposits. The inter-visibility of deposits and trend of formations are perceptibly seen in real 3D geometric environ brought to the study area by x, y values together where z values serves as actual elevations. Here, the maximum perception for interpretation is carried out by the adjustment of sun positioning (azimuth, angle and elevation) and headlight.