ObliqueFit: A MATLAB Tool for 2D Oblique Spatial Analysis Using LiDAR and Photogrammetric Point Clouds

Three-dimensional point clouds acquired through LiDAR and photogrammetric surveys are increasingly used in geoscience and remote sensing workflows. However, generating oblique, georeferenced perspective imagery from these datasets, especially at user-defined viewing geometries, remains a gap in standard processing procedures. Most GIS and 3D analysis platforms (e.g., CloudCompare) do not support direct production of georeferenced oblique rasters, limiting the integration of oblique perspectives into spatial analysis workflows. This limitation arises primarily because GIS systems are designed to work with orthorectified (nadir, distortion-free) imagery.

ObliqueFit, an open-source MATLAB application, is introduced for generating synthetic, georeferenced oblique-perspective images from LAS/LAZ point cloud files for any user-defined camera geometry. The tool provides full control over azimuth, elevation, roll, field of view, and camera distance, as well as quick presets for common viewing configurations. Additionally, ObliqueFit automatically computes the dip and dip direction of the rendered image plane, offering direct structural orientation information tied to the synthetic view. Export options include standard RGB GeoTIFFs, six-band floating-point TIFFs encoding per-pixel RGB and XYZ coordinates, metric depth rasters, and tabulated XYZ coordinate files, all using user-specified EPSG codes via the MATLAB Mapping Toolbox.

Validation with LiDAR and UAV datasets demonstrates geometric consistency between the interactive 3D viewer and exported raster products, particularly for 2D-to-3D projection of geological features. While georeferencing of oblique views is inherently approximate due to non-uniform scale and perspective distortion, the integration of per-pixel depth data (as provided by ObliqueFit) enables accurate 2D-to-3D back-projection. ObliqueFit thus enables accurate, reproducible oblique-perspective analysis from point cloud data, supporting advanced and flexible workflows for remote sensing research and geoscience applications. The tool is freely available under the MIT license on GitHub and archived on Zenodo (https://doi.org/10.5281/zenodo.20458859).