This thesis will analyze innovative classification procedures with the aim of reducing the negative impacts related to the spectral ambiguity and spatial complexity of the land cover classes of VHR images. In order to avoid such disadvantages, the object-oriented classification methodology on Deimos-2 images (1 m GSD PAN and 4 m GSD RGBNir MS) on the study sites of Bari with a high prevalence of plastic roofing for agriculture. The analysis consists of two phases: multi-resolution segmentation and classification of the nearest neighbor of the resulting image objects. With this procedure the generated segments act as image objects whose physical and contextual characteristics can be described using fuzzy logic. Finally, the validity and limits of this methodology for the classification of land cover will be assessed.
This thesis will analyze innovative classification procedures with the aim of reducing the negative impacts related to the spectral ambiguity and spatial complexity of the land cover classes of VHR images. In order to avoid such disadvantages, the object-oriented classification methodology on Deimos-2 images (1 m GSD PAN and 4 m GSD RGBNir MS) on the study sites of Bari with a high prevalence of plastic roofing for agriculture. The analysis consists of two phases: multi-resolution segmentation and classification of the nearest neighbor of the resulting image objects. With this procedure the generated segments act as image objects whose physical and contextual characteristics can be described using fuzzy logic. Finally, the validity and limits of this methodology for the classification of land cover will be assessed.
AN OBIA APPROACH FOR MAPPING PLASTIC GREENHOUSES IN THE BARI AREA USING DEIMOS-2 AND SENTINEL-2 SATELLITE IMAGES
LADISA, CLAUDIO
2020/2021
Abstract
This thesis will analyze innovative classification procedures with the aim of reducing the negative impacts related to the spectral ambiguity and spatial complexity of the land cover classes of VHR images. In order to avoid such disadvantages, the object-oriented classification methodology on Deimos-2 images (1 m GSD PAN and 4 m GSD RGBNir MS) on the study sites of Bari with a high prevalence of plastic roofing for agriculture. The analysis consists of two phases: multi-resolution segmentation and classification of the nearest neighbor of the resulting image objects. With this procedure the generated segments act as image objects whose physical and contextual characteristics can be described using fuzzy logic. Finally, the validity and limits of this methodology for the classification of land cover will be assessed.I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14253/92309