Normalize grayscale for HOG feature extraction

Hello, i'm working on image recognition project. My feature extraction will be using HOG. How important is it to normalize grayscale for HOG feature extraction? Is it enough for using rgb2gray only?

Thank you

Answer 1

[Gautam]

Hello Mark,

Normalizing your images is an important step that can affect the quality of the extracted features. Normalizing the grayscale values ensures that the input to the HOG algorithm is consistent across different images. It also helps mitigate the effects of contrast and illumination differences.

Moreover, HOG is sensitive to edges and gradients, and normalization can help ensure that these features are not overshadowed by variations in lighting.

Using “rgb2gray”, you can convert an RGB image to grayscale, but it does not perform normalization.