Link to Report:

https://drive.google.com/file/d/1JPImtG958uLn4a3SXbMhufUAOKeWhQBi/view?usp=sharing

Dependencies

• Windows 10 64bit with GTX1070
• python==3.5.3
• anaconda==35
• Distance==0.1.3
• (optional)jupyter notebook
• Keras==2.0.4
• matplotlib==1.5.3
• numpy==1.11.2
• openpyxl==2.4.0
• opencv3==3.1.0
• pillow==3.4.2
• scikit-image==0.13.0
• scikit-learn==0.18.1
• scipy==0.18.1
• six==1.10.0
• tensorflow-gpu==1.0.1
• Theano==0.9.0
• tk==8.5.18
• tqdm==4.11.2
• vs2015_runtime==14.0.25123

How to run

1. cd to root directory(the directory that contains this README.md file)

2. You have to clear all contents in result directory

3. run this command :

   python src/launcher.py --original-picture=9.jpg

4. If you like to change the input image, put the image beside this README.md file and change the file name behind --original-picture=

5. Check for the existence of:

   • An empty result directory
   • A FinalResults directory
   • An input image (e.g 9.jpg)

6. Do not delete/modify any unmentioned files/directories

Final Result File (After Execution)

• In FinalResults directory
• There are two files:
  • FinalWordOutPut.txt
    • Text extracted from the given image
    • We give a new line charactor every 10 words
  • FinalOutput.png
    • Modified the font of the original image to ToThePoint.ttf