Engine 2 vs Engine 1

OCR.space Free OCR API
1

Hello. We’re new to the API, we just set up a local server. I’m just trying to see if the behavior we’re seeing is normal.

I have a series of TIFFs that I’m running through Engine 1. Found one that is a particularly low-res scan, that the parser doesn’t correctly decipher the words (even with scale=true). However, if I create a PDF from the TIFF and run it though Engine 2 (on the ocrspace website), it does correctly decipher the text.

Is this normal/expected behavior, or would this be a fluke? Trying to build out the work flow, and also see if we need to purchase Engine 2.

Thanks,
Lori

2

You are correct, usually engine2 is better for English and Western alphabets in general. I use it as default for most of my OCR.

See also