-
Use ‘git clone https://github.com/varunbhardwaj626/Simple_OCR.git ‘or download zip from https://github.com/varunbhardwaj626/Simple_OCR .
-
Unzip the project and locate to the \src in your preferred environment.
-
Install the required python packages if not already(eg -: cv2, numpy, random, PySimpleGUI, tensorflow, os, etc)
-
Run “python main.py” in your environment
Note - If the GUI doesn't show you try installing specified packages or contact the coordinators.
-
The Project can take input of a word at a time due to model limitations.
-
The model is trained on IAM dataset which consists of writing done using thick ink and thereby the model has accuracy based on that.
-
The model has low accuracy over real-world data due to lack of trainable dataset and non-availability of powerful machine required to train more accurate data.