Spatial data mining on remote sensing perspective
International Journal of Latest Trends in Engineering and Technology · 2016
It has been seen that a special interest in the spatial data mining for the past two decades as the technology to collect the aerial images of even remote locations have grown affluently.An abundant store of remote sensing /satellite data is available with the major space research agencies of various nations spread across the continents.Satellite images carry rich spatial information of the coverage area, which includes land use, land cover, vegetation, crops and cultivation, regional population, water bodies, soil, unutilized cultivars, buildings, roads and everything on the surface of the geographical area.The images procured at different time stamps can tell stories of the life systems and its transformations on various perspectives.Therefore, the spatial, temporal and spatial-temporal data mining of satellite images has been emerged as an important branch in data mining research.Similarly, it has become an enthusiastic research area for the researchers of diverse disciplines as it help in decision making directly or indirectly.As these images contain multispectral bands and incorporate several features, they occupy disk spaces of size in the gigabyte terms or more, the processing of images in time bound manner has become a real challenge for the scientists.The usual techniques of classification, clustering and association rule mining for the frequent patterns under the spatial paraphrase can made available to detect and discover unusual patterns and relationships of the objects that are encapsulated in