I scan a lot of bills and I have X images/PDF/JPG files representing the bills.
The program, written in JAVA, has to watch at the folder of scanned bills. When he finds a new file, he has to
1- select the kind of bill by a BAR CODE that is already in the image.
2- choosen automatically the kind of bill he must knows where are the main metadata area (sender, number of protocol, data, total amount)
3- cut the image in several images each one with the sender area, the number area.... and so on
4- run OCR tesseract on that splitted images
5- show me a form with all recognized metadata.
We need the source code in java,
You must use opensource libraries (for example for the barcode choose).