Utilização de imagens omnidirecionais georreferenciadas como controle de campo para orientação de imagens orbitais

• Ano de defesa: 2014
• Autor(a) principal: Olivete, André Luís [UNESP]
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Cartografia; Fotogrametria; Sensoriamento remoto; Georreferenciamento; Cartography
• Link de acesso: http://hdl.handle.net/11449/132243; http://www.athena.biblioteca.unesp.br/exlibris/bd/cathedra/24-11-2015/000854468.pdf

The increasing of the usage of satellite images in the maps and orthoimages production makes the elimination of residuals systematic errors, generated by inaccuracy in the direct determination of the platform motion, fundamental to achieving the necessary precision. These errors can be corrected through control points, but the survey of control points has great cost in the final product, which is related to time and number of professionals required for that activity. The use of an image with an aerial view of the control point, with the precise location of this point, would allow the automatic determination of correspondence between that aerial view and the orbital image, determining the control point position in the orbital image. The aerial view image may be acquired using a catadioptric camera with a conical mirror directed to the ground, and the camera optical axis pointed to this mirror. The georeferencing of catadioptric image may be performed in motion and quickly through the use of a GNSS receiver integrated with inertial measurement unit. This project aims to develop a system for collecting and processing georeferenced catadioptric omnidirectional images as ground control for satellite images orientation. Hardware and software elements have been developed for geometric and radiometric processing of the collected images. Those elements were used to produce scenes with aerial view whose central point has known coordinates and can be located in the orbital images by area based image matching methods. The experiments with ALOSPRISM images showed that correct matches to a sufficient number of points were obtained, allowing geometric image correction