Observatorio de I+D+i UPM

Descripción
Change detection in imagery is quite useful generally but it has particular value in the remote sensing context. The conventional change detection methods based on pixel, which are ap-propriate to the low/middle resolution images, could not model the high-resolution imagery very well. In this case the object-based change detection techniques become effectives. The main ad-vantage of the object-based approaches is that the digital image is no longer considered as a grid of pixels, but as a group of regions, called image objects. In this research, an effective change detec-tion method based on image segmentation is performed. The proposed method involves filtering bands of a multispectral image with a bilateral filter, and obtaining the changed objects by means a method based on a region growing approach. For the analysis of the quality of the resulting seg-mentation, a measure of intrasegment homogeneity (variance indicator) and one of intersegment heterogeneity (Moran index) have been used. The changed segments are placed into the multitem-poral images to identify the nature of change (from-to). In urban expansion areas, the results are promising and show that object-oriented systems facilitate the interpretation of change detection results derived from commercial satellite data.
Internacional	Si
DOI
Edición del Libro	1
Editorial del Libro	Lena Halounova. Czech Technical University in Prague
ISBN	978-80-01-04868-9
Serie
Título del Libro	Remote Sensing and Geoinformation not only for Scientific Cooperation
Desde página	239
Hasta página	246

Esta actividad pertenece a memorias de investigación

• Autor: Javier Sanchez Hernandez (UPM)
• Autor: M.Estibaliz Martinez Izquierdo (UPM)
• Autor: Agueda Arquero Hidalgo (UPM)
• Autor: Diego Renza Torres (UPM)

• Creador: Grupo de Investigación: Informática Aplicada al Procesado de Señal e Imagen
• Departamento: Arquitectura y Tecnología de Sistemas Informáticos