...or so it seems to me.
I am not familiar with any documentation on the subject of OCR effectiveness as it relates to the size of target text.
I just had a multi-page document that I wanted to be reduced (as a way of staying within size constraints of Page Merges) and I found that if I reduced the page sizes to 10% of the original and then ran OCR, the results were hopeless. Of I OCRed and then resized, the text content survived. This is important to me but I have no idea (other than this one experiment) what results to expect. As I now know that reducing before OCRing is a bad idea, I am wondering of resizing bigger would IMPROVE OCR results for problematic OCR runs.
Comments?
Size Matters? It Does When It Comes To OCR
Moderators: TrackerSupp-Daniel, Tracker Support, Paul - Tracker Supp, Vasyl-Tracker Dev Team, Chris - Tracker Supp, Sean - Tracker, Ivan - Tracker Software, Tracker Supp-Stefan
Size Matters? It Does When It Comes To OCR
All best,
FringePhil
FringePhil
- Dimitar - Tracker Supp
- Site Admin
- Posts: 1778
- Joined: Mon Jan 15, 2018 9:01 am
Re: Size Matters? It Does When It Comes To OCR
Hello PHK,
As you well know, quality images are larger in size and OCR relies on image quality to better recognize characters. That's why it's called optical recognition.
It's common sense to optimize a document after you convert it because you won't have one large image to optimize. You'll have smaller images and other components that are much easier to compress.
The formula for success is to use quality (larger) images.
Regards.
As you well know, quality images are larger in size and OCR relies on image quality to better recognize characters. That's why it's called optical recognition.
It's common sense to optimize a document after you convert it because you won't have one large image to optimize. You'll have smaller images and other components that are much easier to compress.
The formula for success is to use quality (larger) images.
Regards.
Re: Size Matters? It Does When It Comes To OCR
Yes, but always being a bit short in the common sense arena it heretofore hadn't occurred to me that resizing the page(s) would give OCRing a better shot. Sorry.
All best,
FringePhil
FringePhil
- Tracker Supp-Stefan
- Site Admin
- Posts: 17824
- Joined: Mon Jan 12, 2009 8:07 am
- Location: London
- Contact: