Türkçe

Research

Visual navigation and geospatial intelligence

My research focuses on visual UAV localization in GNSS-denied and GNSS-degraded environments. I investigate cross-view matching, terrain-aware positioning, remote sensing, photogrammetry and deep learning methods for robust aerial navigation.

  1. Cross-view UAV localization

    Relating observations acquired from an unmanned aerial vehicle to georeferenced aerial or satellite representations in order to estimate position across changing viewpoints.

  2. Terrain-aware visual localization

    Using terrain structure, landform cues and geospatial context as part of visual position estimation rather than treating imagery independently from the landscape.

  3. GNSS-denied and GNSS-degraded navigation

    Positioning approaches for environments where satellite navigation is unavailable, unreliable or insufficient, with emphasis on image and terrain-derived information.

  4. Deep learning for UAV altitude estimation

    Investigating learned visual representations that can support altitude and relative scale estimation from UAV observations.

  5. Orthophoto, DSM, DTM and elevation model integration

    Combining orthophotos, digital surface models, digital terrain models and related elevation products for spatial analysis and navigation research.

  6. Archaeological feature and site detection

    Applying remote sensing and geospatial artificial intelligence to identify landscape patterns that may support archaeological prospection.

  7. UAV photogrammetry and geospatial artificial intelligence

    Connecting photogrammetric products and spatial data with machine learning methods for interpretation, modelling and location-aware applications.