A common approach to stereopsis accomplished using correlation metric to find the closest match in the other image.
Correlation is defined as the product of two normalized vectors divided by their norm.
- Python 3
- Rectified Stereo Images
Test set of Flickr1024, dataset.
Run the following command
python correlation.py