LiDAR Relief Visualization 1.3.3 — QGIS Python Plugins Repository

Changelog

Version 1.3.3 Changelog
Major update bringing core algorithms into strict alignment with
peer-reviewed Relief Visualization Toolbox (RVT) methodologies.

✨ New Features & Algorithmic Upgrades

Enhanced 4-Scale Topographic Position (e4MSTP): Completely rebuilt the
composite engine to use the true 4-step Kokalj (2025) methodology. It
now normalizes and blends Openness (Pos/Neg), Local Dominance, Slope,
dual-scale SVF, and MSTP into a highly nuanced RGB output.
True Horizon-Scanning Local Dominance: Replaced the legacy uniform
filter with a highly accurate radial ray-trace algorithm that
calculates the mean horizon angle arctan(dz/dist), outputting robust
byte-scaled [0-255] local dominance data.
SVF 1D Look-Ahead Noise Filter: Introduced a memory-buffer tracking
system inside the Sky-View Factor (SVF) loop. The algorithm now
successfully identifies and ignores artificial data spikes (noise)
rather than registering them as false terrain peaks.
Research-Validated Terrain Presets: The Batch Visualization tool now
features four static, context-driven presets (Flat/Agricultural,
Forested, Upland/Steep, Coastal) specifically tuned for archaeological
and geomorphological prospecting.
Comprehensive Documentation: Added a detailed USER_GUIDE.md covering
all algorithms, best practices for radii and parameter configuration,
and tips for interpreting composite outputs.
🛠️ UI & Styling Improvements

Granular Batch Controls: The Batch Processing tool now supports
"Manual" overrides, giving users full granular control over search
radii, directions, and observer height.
Universal Auto-Styling: QGIS auto-styling (ReliefLayerPostProcessor)
is now natively applied to all algorithms (including VAT, Simple Red
Relief, and ASVF) as soon as they are loaded into the map canvas.
🐛 Bug Fixes & Stability

Fixed byte-scaling mismatch issues across multiple tools that caused
RuntimeWarning: invalid value encountered in cast errors when
generating uint8 arrays.
Hardened internal algorithms with robust np.nan_to_num() handlers.
Fixed an issue with GitHub Actions CI failing due to missing
scientific testing dependencies (scipy).

Approved

yes

Author

mabo

Uploaded

2026-05-31T06:02:24.276285+00:00

Minimum QGIS version

3.0.0

Maximum QGIS version

4.99.0

External dependencies (PIP install string)

None