HOG

Histogram of oriented gradients (HOG) Python implementation using NumPy from scratch.

HOG is an image feature descripts to describe the image based on the gradients directions and magnitudes. At the current time, this project supports calculating the following:

1. Horizontal and vertical gradients.
2. Gradient magnituge.
3. Gradient direction.
4. Histogram for a given cell.

In the future, the project will support the following:

1. Extracting histogram from all cells within the entire image patch.
2. Concatenating the histogrms of all cells within the same block.
3. Normalize the feature vector of each block.
4. Concatenating all feature vectors from all blocks.

For contacting me:

ahmed.f.gad@gmail.com

ahmed.fawzy@ci.menofia.edu.eg

LinkedIn: https://www.linkedin.com/in/ahmedfgad

Facebook: https://www.facebook.com/ahmed.f.gadd

Twitter: https://twitter.com/ahmedfgad

KDnuggets: https://www.kdnuggets.com/author/ahmed-gad

TowardsDataScience: https://towardsdatascience.com/@ahmedfgad