Skip to content

Repository files navigation

Vision-Based Quality Inspection

Python + Basler camera + ABB robot system for automatic GOOD/BAD part inspection, using a supervised ML classifier and TCP/IP.

Components

  • main.py – Runs online inspection loop and TCP server for the robot.
  • vision_pipeline.py – Basler Camera wrapper and image preprocessing/ROI utilities.
  • classifier_training.py / classifier_evaluation.py – Train and evaluate the image classifier.
  • BaslerCapture*.py, preprocess_images.py, augment_images_in_folder.py, chose_ROI.py – Dataset capture, ROI selection, preprocessing and augmentation tools.
  • abb_robot_comm.py – TCP communication with ABB robot.
  • MesaMainSYI.mod – RAPID program integrating the robot with the vision system.

Usage (online inspection)

  1. Configure IP, camera, and model paths in main.py, abb_robot_comm.py and RAPID Script.
  2. Start the Python server
  3. Start the ABB Script, which connects to the server when available and chosen in setting flags.

Obviously, the model needs to be retrained with a proper dataset that fits the desired usecase.

Video

A simple video is available on Youtube (unlisted), which shows the Robot's movement and realtime evaluation, but does not show the code output. In the video, a good Part is processed and stored on the table.

Demo Video

About

Basler camera + ABB robot vision-based GOOD/BAD part inspection with PyTorch classifier.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages