Recommended scanning settings for pages to be OCRed are the following:
- 300 dpi for regular texts (printed in fonts of size 10 pt or larger)
- 400-600 dpi for texts printed in smaller fonts (9 pt or smaller)
- If you scan very high-quality documents with plain text, then for the best speed use the black-and-white scanning mode.
Note: please make sure you use "threshold" binarization. Dithering, such as "diffusion dithering" or "pattern different" does not fit for OCR.
- If there any garbage, shadows, texture etc. to be removed, use the grayscale mode.
- If any color objects should be removed (for example, stamps), scan the image in the color mode and then use color filtering.
Black and white scanning maximizes scanning speed but may result in the loss of some character information. This may lower recognition quality in documents of medium and low print quality.
A medium brightness value of around 50% should suffice in most cases. During grayscale scanning, brightness is adjusted automatically.