This project enables real-time detection of LEGO bricks in a live camera feed.
Using OpenCV, colors and shapes are segmented, classified, and displayed visually.

• Live video capture from a connected camera
• Color segmentation for LEGO colors: Blue, Green, Red, Yellow
• Shape and size analysis to distinguish different LEGO brick types:
  • 1x3, 1x4, 2x2, 2x4
• ROI extraction (Region of Interest) with bounding boxes
• Console with real-time object overview (incl. position & rotation angle)
• Template matching for more accurate classification (optional)
• Control panel with trackbars to adjust:
  • Refresh rate
  • Display of original frame, color segmentation, channels, and results

├── main.py                # Entry point of the program
├── program.py             # Main loop (camera, segmentation, UI)
├── algorithms.py          # Color and shape algorithms
├── algorithms_tm.py       # Template matching algorithms
├── dataclasses.py         # Definition of LEGO colors, shapes, and data structures
├── fileConverter.py       # Helper functions to convert ROIs
├── consoleWriter.py       # Console output & logging
├── deviceManager.py       # Camera handling
├── imageConverter.py      # Image transformations (HSV, cropping, ROI)
├── ui.py                  # Visualization, bounding boxes, overlays
├── .env                   # Project configuration (e.g. camera settings)

• Python 3.9+
• Installed dependencies (OpenCV, NumPy, python-dotenv)

1. Clone the repository:

   git clone https://github.com/PapeMarc/LegoObjectDetection.git
   cd LegoObjectDetection

2. Create and activate a virtual environment (optional but recommended).
3. Install dependencies:

   pip install -r requirements.txt

4. Create and configure a .env file (example):

   PROJECT_DIR=C:/Users/.../LegoObjectDetection
   CAMERA_WIDTH=640
   CAMERA_HEIGHT=480
   DEF_REFRESH_RATE=200
   MIN_REFRESH_RATE=100
   MAX_REFRESH_RATE=1000
   IMSHOW_SCALE=1
   CAPTURE_NUM=0
   

Run the project with:

python main.py

You will see:

• A Control Panel with trackbars
• Windows for the original frame, color analysis, results, and console

The trackbars allow you to control which outputs are visible and the refresh rate.

• Support for additional LEGO colors and brick types
• Save detection results to files (CSV, JSON)
• Train a machine learning model for improved recognition