Character-Segmentation-of-Scanned-Text
The purpose of this project is to take scanned handwritten/typed documents (image files) as input and generate character level segmentation of its contents.
Model Constrains:
- Document must be written with uppercase and lowercase letters commonly used in English language
- Digits, Signle Numerical Symbols (0-9) can be segmented
Instructions to Execute:
- If Robot Operating System(ROS) is installed in your system then please uncomment the Line 2 and Line 4
- Make sure the installed opencv version must be 3.4.3.18
- If required, modify the input image file name at Line 275
- Command to run the program in Linux/Ubuntu Terminal: python3 main.py
