Heinaunghtet/myNRC-OCR ? reverse-engineered prompt

Build me a simple web app for Myanmar NRC OCR that lets a user upload a card image and then shows the extracted text in a clean, easy to read result view.

I want the flow to feel like a practical KYC tool, so first clean up the image, then detect the important NRC fields, run OCR on each part, and finally apply a bit of text correction so the output looks more human readable. Please make the interface show the uploaded image, the processed image, and the final recognized text side by side or in clear sections, with a loading state while it works and a friendly error if the image is too blurry.

If it helps, use the sample data and existing notebooks as a guide, and look up current docs online if you need to. Keep it lightweight and easy to run locally, and make the main page polished enough for a demo.