Thanks to the very high resolution and the good radiometric quality of the images acquired by GeoEye-1 and WorldView-1, it seems possible to extract a Digital Surface Model (DSM) at accuracy level, comparable to those coming from middle scale aerial products. The accuracy level of DSM is strictly related both to the image orientation and to the matching process. The orientation methods can be classified in two categories: physically-based models (also called “rigorous” models) based on the well-known collinearity equations and generic models that consist in purely analytic functions linking image to terrain coordinates in the form of ratios of polynomial thanks to known Rational Polynomial Coefficients – RPCs. As regards the image matching we can distinguish two basic techniques, the Area Based Matching (ABM) and the Feature Based Matching (FBM). In ABM method a small image window in the master image is opened and on the slave image a template window is shifted and in each position the high correlation coefficient is searched. Whereas FBM searches basic features (like corners, edges, lines, etc) in the two images and after analyzes the correspondence. In this work the accuracy level of DSM extracted with the scientific software SISAR developed by the University of Rome “La Sapienza” is evaluated; the software is able to orientate the high resolution satellite imagery and to generate the RPCs coefficients starting from own rigorous models. The image matching algorithms, implemented in SISAR software, are the FBM and the ABM, the latter is guided by RPC extracted “ad hoc” for the interest area. With this strategy two DSMs are extracted: the first is extracted from WorldView-1 stereopair and covers the Augusta area (Sicily), it is compared with the Lidar ground truth, the latter is extracted from GeoEye-1 stereopair, it covers the Rome area and a 3D vector cartography at 1:2000 scale is used as ground truth for DSM accuracy assessment.
Algoritmi di matching per estrazione di DSM in aree urbane da immagini satellitari ad alta risoluzione / Crespi, Mattia Giovanni; Capaldo, Paola; Fratarcangeli, Francesca; Nascetti, Andrea; Pieralice, Francesca. - ELETTRONICO. - (2010). (Intervento presentato al convegno 14a Conferenza Nazionale ASITA tenutosi a Brescia nel 9-12 novembre 2010).
Algoritmi di matching per estrazione di DSM in aree urbane da immagini satellitari ad alta risoluzione
CRESPI, Mattia Giovanni;CAPALDO, PAOLA;FRATARCANGELI, Francesca;NASCETTI, ANDREA;PIERALICE, FRANCESCA
2010
Abstract
Thanks to the very high resolution and the good radiometric quality of the images acquired by GeoEye-1 and WorldView-1, it seems possible to extract a Digital Surface Model (DSM) at accuracy level, comparable to those coming from middle scale aerial products. The accuracy level of DSM is strictly related both to the image orientation and to the matching process. The orientation methods can be classified in two categories: physically-based models (also called “rigorous” models) based on the well-known collinearity equations and generic models that consist in purely analytic functions linking image to terrain coordinates in the form of ratios of polynomial thanks to known Rational Polynomial Coefficients – RPCs. As regards the image matching we can distinguish two basic techniques, the Area Based Matching (ABM) and the Feature Based Matching (FBM). In ABM method a small image window in the master image is opened and on the slave image a template window is shifted and in each position the high correlation coefficient is searched. Whereas FBM searches basic features (like corners, edges, lines, etc) in the two images and after analyzes the correspondence. In this work the accuracy level of DSM extracted with the scientific software SISAR developed by the University of Rome “La Sapienza” is evaluated; the software is able to orientate the high resolution satellite imagery and to generate the RPCs coefficients starting from own rigorous models. The image matching algorithms, implemented in SISAR software, are the FBM and the ABM, the latter is guided by RPC extracted “ad hoc” for the interest area. With this strategy two DSMs are extracted: the first is extracted from WorldView-1 stereopair and covers the Augusta area (Sicily), it is compared with the Lidar ground truth, the latter is extracted from GeoEye-1 stereopair, it covers the Rome area and a 3D vector cartography at 1:2000 scale is used as ground truth for DSM accuracy assessment.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
