LilScan is designed for Students, Hobbyists, and Makers

Learn the fundamentals of 3D sensing and point cloud processing by building your own 3D scanner to digitize, modify, and 3D print your ideas.

Powered by Raspberry Pi 4

LilScan is designed to run on a Raspberry Pi 4 connected to a Camera Module V2 from the ground up. Based on Pi's fast raw camera access, LilScan uses state-of-the-art machine vision algorithms to reach an accuracy of 0.1mm.

Easy to Setup and Use

The software setup and control are done via an intuitive GUI available for Windows, Mac OS X, and Linux. It allows managing multiple scanners connected to the same local network and seamlessly integrates with other tools for point cloud processing.

Assembled in less than 20 Minutes

An illustrated guide helps to understand the fundamentals of 3D laser scanning and goes through the assembly step by step without requiring any special knowledge or tooling.

Auto-Calibration

Measuring 3D profiles with high accuracy can be challenging with low-cost hardware. Therefore, LilScan uses advanced state-of-the-art auto-calibration routines to correct for any flex or other changes removing the need for expensive rigid setups.

3D Example Scans

Objects scanned with LilScan and post-processed with CloudCompare and Blender.

Drink Can

3D scan of a deformed drink can from 2 different views (2x2 rotations - two for each laser). The total working time for scanning, cleaning, fusion, and meshing is 12 minutes.

Airport Express

Simple 3D scan of an Airport Express with mesh overlay from 2 different views. The total working time for scanning, cleaning, fusion and meshing is 10 minutes.

Bullyland 75595

Complex 3D scan of a dragon (Bullyland 75595) from 4 different views. The working time for scanning, cleaning, fusion and meshing is 45 minutes.

Features & Specifications

Active 3D Triangulation

LilScan is a structured light system projecting a known light pattern onto the scene. Based on the perspective distortion of the pattern, it derives the scene depth.

Accuracy 0.1mm@200mm

In combination with a standard Raspberry Pi camera, the laser line results in an accuracy of up to 0.1mm at a distance of 150mm to the camera.

3D Point Spacing 0.1mm@200mm

The 3D point spacing is up to 0.1mm along and across the laser line at a distance of around 200mm to the scanned object.

Scan Speed

For single 3D measurements, LilScan records the laser and a background image @1Hz. In scan mode, 3D profiles are recorded @30Hz ~ 50k points/sec.

Eye Safe Laser

LilScan uses a 5mW 650nm Class 2M line laser with a 60° line optic, which is considered eye-safe due to the blink refects of a human eye.

Output Format

LilScan stores all 3D data in the PLY format supported by many free programs like CloudCompareor Meshlab.

From Objects to Watertight Meshes

Conversion of a real object into a digital watertight mesh ready to be sculptured or 3D printed.

1 - Digitizing

LilScan digitizes an object's surface by measuring the perspective distortion of a projected laser line. While scanning, a turntable rotates the object so that the laser line can reach most of the object's surface. In addition, a second laser is used during a subsequent scan to increase the coverage further.

2 - Cleaning and Fusion

After a scan has finished, millions of 3D points describe the objects' surface as a digital copy. Tools like CloudCompare or Meshlab can visualize these points and remove outliers generated by laser reflections. They also support the fusion of multiple scans if parts of an object are missing from the scan.

3 - Meshing

3D points do not capture any relationship to neighboring points. Therefore, tools like CloudCompare use robust assumptions about the surface to derive these relations. Finally, the resulting 3D mesh allows to 3D print the object or to use digital sculpturing techniques to modify the digitized object shape.

About

Nowadays, the cost of 3D printers has dropped low enough to be accessible to small companies and individuals. People can easily download, customize, and print products at home, and they often end up costing less than buying them.

However, the same evaluation never happened for 3D sensing technologies, used by major assembly lines for automatization, quality control, or by engineers for reverse engineering. Without the ability to easily measure the 3D world with sub-millimeter accuracy, it is difficult and time-consuming to manufacture any custom parts that fit perfectly.

Using extensive experience in industry-grade 3D laser scanners, LilScan was born as a private project to serve as a learning platform and support the accurate digitization of hardware parts with low-cost hardware at home.