A computer vision project that inspects O-rings in a manufacturing line and provides automated quality assurance by detecting defective and faulty O-rings.
All of this project was done manually. The only thing OpenCV was used for was reading in and displaying the images.
To detect faulty o-rings I used a number of computer vision techniques including:
- Dynamic thresholding
- Binary morphism
- Semantic segmentation
- Image classification
Since this program is using vanilla python, it takes a while to process compared to using OpenCV which is just a wrapper for high performance C++ code.
To run the program on each O-ring, use python main.py
To run the program on a certain O-ring, use python main.py <o-ring-number>
Daniel Moran