Accuracy Assessment of Orthoimagery Generation from Quickbird Satellite Data

The use of satellite data, aimed at retrieving environmental and territorial information, has driven researchers to solve the problem of the geometric rectification in a determined reference system, by seeking suitable planimetric or altimetrical correction techniques. At the same time, recent diffusion of metric and sub-metric data has extended scientific interest towards cartographic updating and orthoimagery production. In the past, such application fields were limited by low spatial and radiometric resolutions for large area investigations. With second generation sensors, panchromatic data are the most suitable for cartographic production, because they permit greater detail and definition of geometric propriety than multispectral data. The most widespread mathematic methods of geometric processing are the parametric and non parametric models. Their use is connected to the availability of external sensor orientation parameters and the possibility to correct optic and terrestrial curvature distortions. In this study QuickBird panchromatic satellite data with sub-metric spatial resolution were processed, in order to assess the accuracy in orthoimagery generation, with the aid of Ground Control Points (GCPs) and of a DTM with a 5 m spatial resolution. Moreover, the results of RMS obtained with consolidated geometric transformation models was evaluated (5th order Polynomial Functions, Rational Function Model, Thin Plate Spline), acquiring GCPs by means of GPS-RTK methodology (with GSM correction).