tthom289

tthom289 / pointcloud-reconstruction

Public

Offline deskewing of VLP-16 point clouds from a rotating LiDAR mount using encoder angles and ICP-based slice alignment, with a first-person OpenGL flythrough viewer for the reconstructed scan.

18
2
100% credibility
Found Mar 17, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Offline toolset to process rotating LiDAR scan data into motion-corrected, aligned 3D point clouds with an interactive viewer.

How It Works

1
🔍 Discover StaticScan

You find a handy tool that turns wobbly spinning laser scans into crystal-clear 3D models of rooms or objects.

2
🛠️ Get your setup ready

You easily add the simple helpers to your computer so everything works smoothly.

3
📁 Load your scan recordings

You point the tool to the folder with your laser scan videos from the spinning platform.

4
✨ Straighten out the blur

The tool automatically fixes motion blur in every frame using the platform's turn angles, making points sharp and true.

5
🧩 Stitch into one big map

It smartly slices and aligns overlapping sections to build a single, perfectly registered 3D point cloud map.

6
👀 Dive into the viewer

You launch the fly-through viewer to zoom, pan, and explore your 3D model like walking through a virtual space.

🎉 Perfect 3D scan ready

Now you have a detailed, accurate 3D model to measure, analyze, or share your scanned environment.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is pointcloud-reconstruction?

This Python tool reconstructs distortion-free 3D point clouds from Velodyne VLP-16 LiDAR scans in ROS 2 MCAP bags recorded on rotating platforms. It deskews each frame using encoder angles, aligns angular slices via ICP for drift-free maps, and outputs PCD files—no ROS install needed, even decompresses zstd bags automatically. A first-person OpenGL viewer lets you fly through results, compare scans, and tweak alignments interactively.

Why is it gaining traction?

Unlike online SLAM stacks, it runs fully offline on bags, sidestepping real-time dependencies while handling mechanical imperfections better than pure deskewing. The viewer shines with grab-drag positioning, pick-based ICP refinement, and outlier filtering, making point cloud reconstruction github workflows visual and forgiving. Customizable params like rotation axis and mount offsets adapt to custom hardware without recompiles.

Who should use this?

Robotics engineers processing rotating LiDAR bags for mapping or inspection, SLAM devs validating offline point cloud surface reconstruction from VLP-16 data, or surveyors turning encoder-synced scans into aligned PCDs. Ideal if you're debugging alignments manually or comparing deskewed vs ICP-merged clouds.

Verdict

Grab it for niche rotating LiDAR setups—docs are thorough, CLI is dead simple, and the viewer alone justifies a test run. With 18 stars and 1.0% credibility score, it's early-stage and unproven at scale, but mature enough for prototypes if your bags match the topics.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.