Convert a ROS1 LiDAR and IMU bag file into a 3D point cloud map without setting up Linux or ROS.
Preview a scan's tracking quality and trajectory live while it processes.
Export a completed map as PLY or PCD for use in other 3D or robotics tools.
Test mapping quality against public LiDAR datasets like the TIERS multi-modal dataset.
| musabali314/onslam | 0-bingwu-0/live-interpreter | 0cm-labs/tokenizer-benchmark | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | Python | Python | Python |
| Setup difficulty | — | moderate | moderate |
| Complexity | — | 2/5 | 2/5 |
| Audience | developer | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
OnSLAM is a Windows desktop application that turns a recorded LiDAR and IMU data file from the ROS1 robotics format into a navigable 3D point cloud map, without requiring the user to install Linux, ROS, Python, or Node.js, or type any terminal commands. Everything runs locally on the user's own computer, so no bag files, maps, or logs are ever sent anywhere else. After installing it, a user opens a recorded .bag file inside the app, which shows the available LiDAR and IMU data streams inside that file. They pick which stream to use, choose a processing quality preset such as fast, balanced, or detailed, and start the mapping process. The app then combines the LiDAR and motion sensor data using an IMU-assisted technique that matches each new scan against a growing local map, shows live progress including the estimated path and a tracking quality indicator, and previews the point cloud building up in the browser as it goes. Once finished, the result can be downloaded as a PLY or PCD file, two common formats for storing 3D point cloud data. Under the hood, the installer bundles a Python and FastAPI backend that does the actual processing, a compiled React interface for the browser view, and the various supporting libraries, and launches a local web service that opens automatically at an address only reachable from the same machine. It currently supports two types of LiDAR data (Livox's own format and the more generic PointCloud2 format) paired with standard IMU data, but does not yet support ROS2 bag files, correcting for accumulated drift over a full loop, or automatically figuring out how sensors are physically positioned relative to each other. The developer describes it as an experimental and practical tool rather than a full replacement for more complete mapping systems used in robotics research. Processing large recordings can use a lot of memory, CPU time, and temporary disk space depending on the length of the recording and the chosen quality setting. The README does not state a license for the project.
A Windows app that turns a ROS1 LiDAR and IMU recording into a downloadable 3D point cloud map, all processed locally.
Mainly Python. The stack also includes Python, FastAPI, React.
Mainly developer.
This repo across BitVibe Labs
double-check against the repo, no cap.