Lakin Wecker, PhD
Writing
- Constructing Efficient Mesh-Based Global Grid Systems with Reduced Distortions
Ph.D. thesis. Introduces a mesh-based Discrete Global Grid System (MBD) that achieves both low distortion and constant-time operations on high-resolution spherical meshes, using novel hash encoding and barycentric indexing. This work is the foundation of bg-core.
- Constructing Efficient Mesh-Based Global Grid Systems with Reduced Distortions
Generalizes efficient DGGS operations over high-resolution spherical meshes, allowing for significantly reduced distortion while maintaining constant-time operations.
- A Multi-Resolution Approach to Point Cloud Registration without Control Points
Georeferences terrestrial photographic imagery without requiring in-situ control points, making structure-from-motion-based monitoring practical in remote and rough terrain.
- Cloud Shadow Detection via Ray Casting with Probability Analysis Refinement Using Sentinel-2 Satellite Data
Uses ray-casting and inverse texture mapping to detect and remove cloud shadows from satellite imagery, improving the reliability of precision agriculture analysis.
- Efficient Calculation of Distance Transform on Discrete Global Grid Systems
A distance transform algorithm for Discrete Global Grid Systems that exploits the hierarchy of a DGGS to compute geodesic distances efficiently and with minimal distortion. Useful for applications like wildfire management.
- Perspective Charts in a Multi-Foci Globe-Based Visualization of COVID-19 Data
Integrates perspective projection into geospatial charts to help readers perceive small values in datasets with variations at multiple scales, evaluated against traditional charts using global COVID-19 data.
- Disdyakis Triacontahedron DGGS
Introduces a Discrete Global Grid System based on the Disdyakis Triacontahedron, a Catalan solid with 120 faces that more closely approximates a sphere and reduces distortion in geospatial data representation.
- A multiresolution approach to iris synthesis
A method for expanding iris image databases through multiresolution decomposition and recombination, making it possible to test iris identification algorithms against larger, more diverse synthetic datasets.
- Synthesizing techniques based on multiresolution
M.Sc. thesis. Explores multiresolution techniques for synthesizing realistic geometric and image data, with applications in iris biometrics and terrain modelling.
- Contextual void patching for digital elevation models
A multiresolution method for repairing voids in digital elevation models by borrowing shape characteristics from surrounding terrain, producing smooth and realistic patches.
- Iris synthesis: a reverse subdivision application
A reverse subdivision method for synthesizing new iris images from a small set of existing ones, aimed at expanding limited iris databases for biometric research.