/sift-hog-image-feature-descriptors

applying SIFT and HOG to localize cropped pieces of text

Primary LanguageJupyter Notebook

Applying SIFT and HOG to localize cropped pieces of text

there are two pipeline functions separately for SIFT and HOG feature descriptors to match a given input image with a specified input hyperparameter to the original text.
The input cropped images may vary in brightness, rotation, scale, and resolution.