How does grid extracts feature descriptors in the Bag of Visual Words?

Accroding to here:

https://www.mathworks.com/help/vision/ug/image-classification-with-bag-of-visual-words.html

If you choose to grid for extracting features from images, how does that work? Apparently it does different than feature detection which is based on speeded up robust features (or SURF).

Answer 1

[Jalaj Gambhir]

Hi,

'grid' algorithm is a Point Selection Method which is the followed by the step of Feature Extraction from the selected points (grids). You can read more about it here ('PointSelection' section and the following 'GridStep' section).

[Zeynab Mousavikhamene]

Any thought about this questin: https://www.mathworks.com/matlabcentral/answers/513266-bagoffeatures-histogram-yaxis-is-decimal would be appreciated.