Voormansik K.,University of Tartu |
Praks J.,Aalto University |
Antropov O.,VTT Technical Research Center of Finland |
Jagomagi J.,Regio Ltd. |
Zalite K.,Tartu Observatory
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing | Year: 2014
In this study, an extensive flood in Estonia during spring 2010 was mapped with TerraSAR-X data acquired over both open and forested areas. This was the first time when a large scale flooding area was mapped in Estonia by means of spaceborne remote sensing. This was also the first time when X-band SAR images were successfully used for flood mapping under the forest canopy in the temperate forest zone. The tree height in the study region was 15-25 meters on average, and main tree species were birch (leaf-off condition), pine and spruce. The results were compared with ALOS PALSAR and Envisat ASAR images of the same flooding event. In the study area, TerraSAR-X provided on average 3.2 dB higher backscatter over mixed forest flooded areas compared to non-flooded areas. In deciduous and coniferous forests the difference in average backscatter between flooded and non-flooded forests was even greater, 6.2 dB and 4.0 dB, respectively. A supervised classification algorithm was developed to produce high resolution maps of the flooded area from the TerraSAR-X images to demonstrate the flood mapping capability at X-band. Our results show, that spaceborne X-band SAR data, which currently has the highest resolution among the SAR instruments in space, can be used to map floods under forest canopy in temperate zone despite its short wavelength and high attenuation. © 2008-2012 IEEE.
Haav H.-M.,Tallinn University of Technology |
Kaljuvee A.,Regio Ltd. |
Luts M.,Competence Center in Electronics |
Vajakas T.,Regio Ltd.
Frontiers in Artificial Intelligence and Applications | Year: 2011
Personalization is very important feature of any Location Base Service (LBS) as it improves its usability. The paper focuses a problem of development of personalized LBSs. In the paper, a novel approach based on ontology engineering is proposed to provide an intelligent (semantics-based) solution to personalization problem of LBS. The approach uses geospatial ontologies, ontology-based user profiling and multilingual output for personalization of services. The provided ontology-driven development framework for personalized LBS is evaluated by implementing the personalized reverse geo-coding service use case. © 2011 The authors and IOS Press. All rights reserved.