My research interests focus on basic research on spatial data, applying remote sensing and Geographic Information Systems (GIS) to solve environmental issues, generating decision support systems for policy makers, web-based image presentation and geostatistics.
In terms of basic research, I have been working on data accuracy, uncertainty, geometric and thematic accuracy, creation of Digital Elevation Models and application of data fusion, pixel and object based image classification techniques to improve the information content extracted from satellite images. On the other hand, I have been addressing the environmental issues like land use/cover changes, water quality modeling and monitoring, climate, agriculture abd soil salinity mapping
I have been using multi-temporal remotely sensed data to detect 2D and/or 3D changes on coastlines, water resources, disaster, urban, agricultural and forest areas.
In terms of image processing, my interests are image-super resolution, pan-sharpening and deep learning based object identification and mapping.
Bu çalışma kapsamında, 150 km2 lik pilot bir alan için, farklı mekânsal çözünürlükteki uydu görüntüleri (30 cm-10 m arasında değişen) nesne tabanlı sınıflandırma yaklaşımı ile sınıflandırılıp arazi örtüsü/kullanımı haritaları oluşturulacaktır. Sonraki aşamada farklı arazi örtüsü/kullanımı haritalarından farklı peyzaj metrikleri hesaplanıp; mekânsal çözünürlük ve sınıflandırma doruluğunun peyzaj metriği hesabına etkisi araştırılacak ve şehir alanların yüksek doğrulukla sınıflandırılması ve peyzaj metrikleri ile analizinde en optimum uydu verisi, peyzaj metrikleri ve çözünürlük-peyzaj metriği ilişkisi hakkında değerlendirme yapılacaktır.
Mapping of the Earth’s land cover and use is essential to understand the natural and manmade processes effecting our environment and conduct future projections for several phenomena such as urbanization, deforestation, hydrological processes, climate change etc. In this study, a novel National Land Cover/Use (NLC) nomenclature, which was created from Coordination of Information on Environment (CORINE) land cover/use classification system by expanding Level 3 CORINE classes to a fourth level including 75 subclasses, is presented and applied to 3 different study areas having different landscape characteristics
TARBIL is a high technologic project aiming to create up-to-date agricultural parcel map of the country, incorporate agricultural parcels with cadastral information, measure the crop yield by commodity and estimate the crop yield for upcoming season. The project aims to estimate the crop yield using the integration of ground and space based measurements.
Overall system will integrate data from different sources and create agricultural information by means of a decision support system which will be benefitted for agricultural policies and management.
In this project, it has been aimed to conduct a pilot research to determine the spatial distribution of vineyard areas in Tekirdag city, to identify the grape diversity in Sarköy district, to create a spectral library for different grape types with respect to phenological periods and to create a national viticulture information system. Vegetation index maps and land cover maps created from Hyperion, CHRIS-Proba, SPOT 5, IKONOS and Worldview-2 satellite images were analyzed to determine the boundaries, spatial and textural distributions of vineyard areas in Tekirdag city.