What do you need help with?
Splitting an image into pieces?
Running OCR (Tesseract) on am Image?
The full code i can not share as it is part of my Discord-Bot (30k lines)
Also it does not help you with your issue.
You have to KNOW the correct position of something to crop an Image based on the bounds you already know.
Look at the image i posted in #3. The positions of the driver, Car, Times are all at a specific positions. On every result-image.