This is our secret sauce. Instead of a single ray-tracing model, we have sensor-specific ray factories .
Sen3DKol was built as a synthesis of robust geospatial libraries (GDAL, PROJ) and high-performance computing algorithms (C++ matching cores). Its architecture prioritizes the specific geometric challenges of the Sentinel-3 mission, providing a streamlined tool for extracting 3D information from a sensor primarily designed for 2D observation. The software stands as a specialized utility for glaciologists, hydrologists, and terrain analysts requiring rapid topographic mapping. how sen3dkol software built
Sen3dKol is not finished. It will never be finished. Every time a new sensor comes out (solid-state LiDAR, FMCW radar, neuromorphic cameras), we have to go back to the physics drawing board. But that’s the beauty of it – we are not building a product. We are building a continuously evolving simulation of reality . This is our secret sauce
The software was constructed following a . This approach was chosen to balance the need for high-performance processing of large raster datasets with the flexibility to swap out specific algorithms (e.g., changing the matching algorithm without rewriting the input/output handlers). It will never be finished
The user interface was constructed using . This choice provides a cross-platform compatible interface (Windows/Linux) that feels native to the operating system. The UI design follows a "Wizard" pattern, guiding the user through:
This is our secret sauce. Instead of a single ray-tracing model, we have sensor-specific ray factories .
Sen3DKol was built as a synthesis of robust geospatial libraries (GDAL, PROJ) and high-performance computing algorithms (C++ matching cores). Its architecture prioritizes the specific geometric challenges of the Sentinel-3 mission, providing a streamlined tool for extracting 3D information from a sensor primarily designed for 2D observation. The software stands as a specialized utility for glaciologists, hydrologists, and terrain analysts requiring rapid topographic mapping.
Sen3dKol is not finished. It will never be finished. Every time a new sensor comes out (solid-state LiDAR, FMCW radar, neuromorphic cameras), we have to go back to the physics drawing board. But that’s the beauty of it – we are not building a product. We are building a continuously evolving simulation of reality .
The software was constructed following a . This approach was chosen to balance the need for high-performance processing of large raster datasets with the flexibility to swap out specific algorithms (e.g., changing the matching algorithm without rewriting the input/output handlers).
The user interface was constructed using . This choice provides a cross-platform compatible interface (Windows/Linux) that feels native to the operating system. The UI design follows a "Wizard" pattern, guiding the user through: