SwiftOCR FAQs

If your question is not answered here you can post a question to the SwiftOCR Support Forum and one of our developers will answer it directly.

1. Does SwiftOCR handle full-page OCR?

No, SwiftOCR is designed to be used in applications regarding fixed fields of text such as Passports and ISBN.

2. Can SwiftOCR recognize any fonts?

No, SwiftOCR currently works with OCR-A, OCR-B and MICR13 fonts.

3. Does SwiftOCR work with handwritten text?

No, SwiftOCR works only on certain standardized machine readable fonts.

4. How do I tell SwiftOCR what kind of text to read?

Templates are used to tell SwiftOCR how wide text fields are, how many rows should be read, what the font should be and what possible character values are allowed at each template location.

5. Do all templates have to be a fixed width?

No, SwiftOCR allows templates of variable length to be defined.

6. What size OCR fonts does SwiftOCR work with.

The font size on the surface to be read is not as important as the size of the imaged OCR character in memory. SwiftOCR will work with imaged characters ranging in height from 10 to 50 pixels, with 20 pixels being the optimal size based on read rate, accuracy and performance.

7. Does SwiftOCR read all characters in OCR-A and OCR-B?

No, but SwiftOCR reads all alphanumeric chars (uppercase A-Z and digits 0-9) as well as many standard punctuation characters.

8. Does SwiftOCR handle mirror imaged text?

Yes, via a SwiftOCR property.

9. Are there any predefined templates provided?

Yes, standard templates to read Passports, ISBN, Price Fields and MICR E-13B are included.

10. Does SwiftOCR provide checksum cababilites within the OCR text?

Yes, several different types of checksums (row based, block based) may be embedded withing the OCR text and then verified before the OCR result is issued.