7 January 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Surface refinement with shading information is a popular approach to recover fine details from the initial surfaces reconstructed from multi-view stereo (MVS) methods. However, the existence of discontinuity areas challenges these methods in urban scenes. To address this need, this paper proposes an edge constrained DSM refinement method based on shading from high resolution multi-view satellite images. The proposed method combines the advantages of stereo matching and shape from shading techniques. First, an improved Semi-Global Matching (SGM) method is applied to reconstruct the initial DSM; then, the initial DSM and the original images are used to detect the discontinuity areas in the scene; finally, the initial DSM is refined by minimizing an edge constrained energy function which models the imaging process with the illumination, surface albedo, and surface normal. Experiments with Pleiades and GaoFen-7 satellite images show that the proposed method can recover the detailed geometry structures well and achieve an improvement of elevation accuracy up to 18.4%, 12.95%, 56.94%, 59.02%, compared to the initial DSM, the recently published shadingbased DSM reconstruction method (SADGE) and two commercial software solutions (SURE and Metashape) respectively
26 May 2025 in Neurocomputing
3 March 2026 in IEEE Transactions on Geoscience and Remote Sensing
Tower-based surveillance (TBS) systems provide continuous, high-resolution, and real-time monitoring over wide spatial extents. However, accurate geolocation of TBS imagery remains largely underexplored due to large viewpoint discrepancies, nonlinear radiometric distortions, and limited spatial overlap with satellite imagery. To address these challenges, we propose TowerSatLoc, a high-precision and fully automated correspondence-based geolocation framework that registers TBS imagery to geo-referenced satellite images under complex conditions. The framework first applies a bird’s eye view (BEV) transformation using intrinsic and extrinsic camera parameters to reduce viewpoint discrepancies. Within this process, an adaptive projection factor (APF) and an automatic orientation estimation method are introduced to improve geometric consistency. To handle severe cross-domain differences, semantic representations extracted from DINOv2 are fused with multiscale geometric features from VGG19 and refined through hierarchical residual matching and confidence-aware resampling. The framework further incorporates robustness-enhancing mechanisms to improve localization reliability under extreme viewpoint variations and uncertain tower geocoordinates. Extensive experiments on a newly constructed dataset, TowSatData, demonstrate the effectiveness and robustness of TowerSatLoc across diverse real-world scenarios, providing a scalable and generalizable solution for correspondence-driven cross-view geolocation.