I'd be happy to help with Tesseract wisdom. I mainly use it to extract text from WhatsApp conversations. Locally, it works perfectly, extracting all the text without any issues. However, the problem arises when I upload it to the cloud with DJANGO. I've configured everything necessary for Tesseract to work. Locally, I manually installed it because I'm using Windows, and for the server, I simply used Docker. Somehow, I encounter version differences between local and production.
But the main issue is that in the cloud, it misses a lot of text whereas locally, everything comes out perfectly. I tried changing the image to send it from React via API to the cloud and download it to my server, but it remains the same. I really don't know why it sometimes misses text and sometimes works perfectly, even in the cloud, this varies from image to image. Needless to say, I also use CV2 to improve its performance significantly. Would appreciate any help on this matter.
Prompt: What steps have you taken to troubleshoot the Tesseract issue both locally and on the cloud? What are the specific differences you've observed between the successful and unsuccessful cases?