OCR Configuration

When setting up a new optical recognition program you need to enter the Config details before adding the details.

General

1. Start by entering a Configuration ID.

2. Add a short description to be used to explain what the OCR is or for.

Ocr Engine

1. Select from the Ocr Engine Type dropdown the relevant type.

2. You can limit the characters recognised by entering them into Allow List Characters.

   • You must include a space-character to retain whitespace.

   • If left blank all characters will be included.

3. You can limit the characters recognised by entering them any into the Deny List Characters.

   • If left blank all characters will be included.

4. Choose the mode the engine will run in from the Ocr Engine Mode.

   • Tesseract Only - Only the legacy Tesseract OCR Engine is used (Tesseract 3 OEM mode).

   • LSTM Only - Only the new LSTM based OCR engine is used. Uses a pre programmed neural net (Tesseract 4 + 5 mode).

   • Tesseract and LSTM - Both the legacy and new LSTM based OCR engines are used (recommended for a good balance of speed and performance).

   • Default - Use the OCR engine currently recommended by the IronOCR developers.

5. You can choose which Tesseract Version from the dropdown.

6. Choose the Language Mode from the dropdown.

   • English - The Default mode, it uses both an LSTM and OEM strategy to produce fast and accurate results.

   • English (Best) - Uses an LSTM engine optimised for detail over speed.

   • English (Fast) - Uses an LSTM tesseract engine optimised for speed over accuracy.

Agent

1. Select the Source Agent you want to use for this Source Configuration.