If I use engine 3 I get less results when I turn on the overlay compared to when it is off.
I ran it on the same set of 12 images and with overlay off I get text results on 8 of the images. With overlay on I get results on only 1 of the images.
This is the set of POST parameters I am using:
apikey: …
base64Image=…
language=eng
OCREngine=3
isOverlayRequired=false
I am using a paid key. I would like to see the text results still returned
Thank you for reporting this issue. Unlike Engine 2, Engine 3 is LLM-based, and it could indeed be that requesting overlay data influences the OCR quality (it should not be, but could be. We will be testing this further.).
In our Engine3 specs paragraph, we mention that we recommend Engine 2 for accurate overlay data:
Engine 3 prioritizes OCR accuracy and markdown output over spatial precision. Overlay data (x/y bounding boxes per word) is available, but coordinates are not as precise as Engine 1/2. Requesting overlay data (isOverlayRequired=true) also makes the call 2-3× slower; on Engine 1/2 the same option has almost no performance cost.
Having said that, we are working on an update that combines the OCR quality of Engine 3 while providing the same spatial accuracy (overlay) as Engine 2.
Thanks. The LLM result makes sense. That would explain why one of my results was this LLM like description of the image. This does not effect me, but just passing this on for feedback:
,"ParsedText":"",