I am new to computer vision. I am trying to extract text from video frames and images. Most of the codes provided in github are compatible with python <3 versions. Any idea on how to proceed and get the related codes and good papers.
Note: I have already implemented pytesseract-OCR and I haven't gotten good result.
Hello TISHANT CHANDRAKAR.
At first, you must understand how text recognizer works.
1. have-text-region is extracted from the image
2. we recognize text for each region
3. Combine the text of all regions to form final result
Tesseract itself work very well for step 2. But for step 1, it only work well with text in document. In computer vision, step 1 is called "Scene text detection". So your next step is find some good code, or paper which could do "scene text detection".
If you want to learn and read paper there is a list here Scene text detection list . But in my opinion, the text in you image is white text on black background so a simple color threshold could solve step 1 easily.
Hope that help