This repository contains a Python script to visualize LiDAR point cloud data projected onto a camera image. The project is part of the "Software Development Methods" lecture in the Advanced Driver Assistance Systems (ADAS) Master's program (2025).

The script performs the following tasks:

• Reads camera image and LiDAR point cloud
• Applies extrinsic and intrinsic camera calibration
• Projects 3D LiDAR points into the 2D image plane
• Colors projected points based on LiDAR intensity
• Displays the resulting image overlay

This visualization helps understand the sensor fusion pipeline, particularly the spatial alignment between LiDAR and camera systems.

├── main.py               # Entry point: performs projection and visualization
├── image_reader.py       # Utility to read and preprocess image files
├── lidar_reader.py       # Utility to load and parse PCD files
├── data/
│   ├── 018282150.png     # Sample image
│   └── 018282150.pcd     # Corresponding LiDAR point cloud

Make sure you have the following installed:

• Python 3.8+
• NumPy
• Matplotlib
• Pandas

You can install dependencies using:

pip install numpy matplotlib pandas

or alternatively

pip install -r requirements.txt

To run the script:

python main.py

Make sure the data directory contains the corresponding .png and .pcd files with matching filenames.

• Camera Intrinsics:

  [[307.43, 0.00, 387.17],
   [0.00, 304.43, 157.75],
   [0.00, 0.00, 1.00]]
  

• Extrinsic Transformation (LiDAR → Camera):

  • Rotation and Translation are encoded in a 4×4 transformation matrix.
  • This matrix aligns the LiDAR coordinate system to the camera frame.

The given point cloud is given as:

Whereby the input image is already undistorted:

The result is a 2D camera image overlaid with LiDAR points color-coded by intensity:

• High intensity → brighter colors
• Low intensity → darker colors

Example output:

This project is intended for educational use within the ADAS Master's program. It is provided along with the MIT license.

Happy coding!