Using Otsu's thresholding for text segmentation on images of sticky notes.
Assumptions
- The image has a dark background below the paper
- The writing is darker than the paper.
- Even lighting (Since global thresholding is used)
- The image is 1:1 aspect ratio
Proof of concept python notebook can be found in html format here
Example 1
Example 2