In the last years, the extraction of the information content from digital images has assumed a crucial role in many application fields, such as risk assessment, analysis of damages, deforestation, environmental monitoring, earth observation, as a fundamental instrument to carry out specific and pointed studies. In this context the change detection of remotely sensed images takes place. Detecting changes means performing a spatial comparison of two or more images acquired over the same geographical area at different times. This operation can be performed on a per-pixel basis as well as on a per-object basis, depending on the aim of the specific application. In particular, in this paper two versions of the same change detection algorithm are presented, the one working on a per-pixel basis while the other working on an per-object basis, applied specifically for the monitoring of a water supply infrastructure. This algorithm provides the changes occurred in optical images' spectral content, as well as in their radiance content, by calculating two change features: the spectral angle made by two corresponding spectral vectors in the compared images, and the so-called Brightness Change Factor. The object-based version of the presented change detection algorithm has been developed according to an IIM - Image Information Mining context, in order to introduce an automated procedure to detect changes; furthermore, it has been developed with an a image analysis framework, called IPAINT - Image Processing Analysis Interpretation and Trasconding, which can be used for various applications thanks to its versatility, since it offers many different techniques for image processing.
A water supply infrastructures application of change detection by measuring spectral change features
D'ELIA, Ciro;RUSCINO, Simona;DE MARINIS, Giovanni;LEOPARDI, Angelo
2012-01-01
Abstract
In the last years, the extraction of the information content from digital images has assumed a crucial role in many application fields, such as risk assessment, analysis of damages, deforestation, environmental monitoring, earth observation, as a fundamental instrument to carry out specific and pointed studies. In this context the change detection of remotely sensed images takes place. Detecting changes means performing a spatial comparison of two or more images acquired over the same geographical area at different times. This operation can be performed on a per-pixel basis as well as on a per-object basis, depending on the aim of the specific application. In particular, in this paper two versions of the same change detection algorithm are presented, the one working on a per-pixel basis while the other working on an per-object basis, applied specifically for the monitoring of a water supply infrastructure. This algorithm provides the changes occurred in optical images' spectral content, as well as in their radiance content, by calculating two change features: the spectral angle made by two corresponding spectral vectors in the compared images, and the so-called Brightness Change Factor. The object-based version of the presented change detection algorithm has been developed according to an IIM - Image Information Mining context, in order to introduce an automated procedure to detect changes; furthermore, it has been developed with an a image analysis framework, called IPAINT - Image Processing Analysis Interpretation and Trasconding, which can be used for various applications thanks to its versatility, since it offers many different techniques for image processing.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
