Skip to content

lidar-strip-adjust

Automated boresight calibration and strip adjustment for airborne and UAV LiDAR point clouds.

PyPI CI License: MIT

What it does

lidar-strip-adjust estimates a 6-DOF rigid correction (Δω, Δφ, Δκ, Δx, Δy, Δz) between overlapping flight strips using a planar-neighbourhood point-to-plane ICP algorithm:

  1. Extracts planar patches from each strip via k-NN PCA + Weinmann planarity filter
  2. Matches patch pairs across strips by proximity and normal agreement
  3. Minimises point-to-plane distances with Levenberg–Marquardt optimisation
  4. Applies the estimated boresight correction and reports RMSE before/after

Installation

pip install lidar-strip-adjust

Requires Python ≥ 3.10.

Quickstart

CLI

lidar-strip-adjust reference.las target.las -o corrected.las

Python API

from lidar_strip_adjust import StripAdjuster, load_strip, save_strip

ref = load_strip("reference.las")
tgt = load_strip("target.las")

adjuster = StripAdjuster()
result = adjuster.adjust(ref, tgt, strip_id="strip_02")

print(f"RMSE before: {result.rmse_before:.4f} m")
print(f"RMSE after:  {result.rmse_after:.4f} m")
print(f"Δω={result.params[0]*57.3:.4f}° Δφ={result.params[1]*57.3:.4f}° Δκ={result.params[2]*57.3:.4f}°")

save_strip(result.corrected_points, "target.las", "corrected.las")

Why it exists

No maintained open-source Python package existed for LiDAR boresight calibration as of early 2026. Practitioners relied on commercial tools (TerraScan, RiPROCESS) or wrote one-off scripts. See the Architecture Decision Record for algorithm rationale and rejected alternatives.

Benchmark

On a 200,000-point synthetic dataset (8-core laptop, Python 3.12):

Metric Value
Wall time ~5 s
Peak memory ~180 MB
RMSE reduction 91 %