Descripción
|
|
---|---|
Change detection is the process of identifying differences in the state of an object or phenomenon by observing it at different times. Most existing multi-temporal change detection methods use the spectral information alone. However, the inclusion of spatial information in change detection increase the accuracy of this process. For other hand, if accurate registration between images were not achieved, false differences are evaluated instead of real changes at the same location. In this work, the authors propose a novel methodology in order to detecte the spectral and morphological changes into a temporal sequence of images over the same geographical area. This methodology is based on arithmetic operations between fractal maps, calculated by the box-counting algorithm, of the images under analyzes. Two multispectral Spot5 images has been used to evaluate the performances of the proposed methodology and the results obtained have been compared with widely used methods. From the results it can be concluded that the new method presents two main advantages against the other method, that increase the changes detection accuracy: the capability of multi-scale change detection and low sensitivity to misregistration. | |
Internacional
|
Si |
DOI
|
10.3233/978-1-60750-494-8-241 |
Edición del Libro
|
1 |
Editorial del Libro
|
European Association of Remote Sensing Laboratories |
ISBN
|
978-1-60750-493-1 |
Serie
|
|
Título del Libro
|
Imagine Europe |
Desde página
|
242 |
Hasta página
|
248 |