Land cover exerts a large influence on many basic environmental processes and consequently any transformation in it may have a marked impact on the environment from local to global scales. In multidisciplinary research contexts satellite remote sensing offers opportunities both to evaluate the effects of these processes and to provide one of the information layers needed for designing national strategies oriented to protection and sustainable use of our resources. The advent of recent satellite imagery has increased the possibility to investigate large scale area with a great level of detail. Associated with an increase in spatial and radiometric resolution there is, usually, an increase in variability within land parcels, generating a decrease in accuracy of land use classification on a per-pixel basis. In order to avoid such negative impacts, an object-oriented classification methodology on IKONOS multispectral data is implemented on the test area of the Alta Murgia National Park, in the Apulia region (Italy), where the soil adaptation to agriculture practices, through rocks breaking, in the last twenty years has occurred. The analysis is conducted with a classification strategy, that is able to distinguish land use functions on the basis of differences in spatial distribution and pattern of land cover forms. It consists in two phases: a first segmentation of the image into meaningful multipixel objects of various sizes, based on both spectral and spatial characteristics of groups of pixels; then, the segments (objects) are assigned to the classes using fuzzy logic and a hierarchical decision key.
|Autori interni:||TARANTINO, Eufemia|
|Titolo:||Spatial Information Extraction From VHR Satellite Data to detect Land Cover Transformations|
|Titolo del libro:||INTERNATIONAL ARCHIVES OF THE PHOTOGRAMMETRY, REMOTE SENSING AND SPATIAL INFORMATION SCIENCES|
|Data di pubblicazione:||2004|
|Appare nelle tipologie:||2.1 Contributo in volume (Capitolo o Saggio)|