Skip to content

Latest commit

 

History

History
38 lines (31 loc) · 1.2 KB

README.md

File metadata and controls

38 lines (31 loc) · 1.2 KB
Raw

OCR-and-Keyword-extraction-using-EasyOCR-and-Yake

Unsupervised learning to extract text from images and determine keywords within the text

Topic

In this project I will experiment with Optical Character Recognition or OCR to convert written information contained within an image to text string containing the same information in the image. I will do this with the help of EasyOCR which is a prebuilt and pretrained network consisting of convolution and recurrent layers. I will then try to extract the keywords present in the extracted text and plot those keywords using WordCloud which plots the words' size proportional to the importance of the word in the text. So let's get started !

Objectives

-Perform OCR to extract text from images

  • Select and plot the most important keywords contained within the text

Summary

  • Importing libraries
  • Case I:
    • Plotting the image
    • Extracting text
    • Extracting an plotting keywords
  • Case II:
    • Plotting the image
    • Extracting text
    • Extracting an plotting keywords
  • Case III:
    • Plotting the image
    • Extracting text
    • Extracting an plotting keywords
  • Conclusion

Libraries

  • EasyOCR
  • Yake
  • CV2
  • Matplotlib

Data source

CNN NEWS