Font recognition

  • Last Post 11 July 2018
  • Topic Is Solved
AdrianHB posted this 03 July 2018

Hi, I'm new here, helping with an installation run by someone else.

I believe in some cases FineReader XML output identifies the font used.  I know it's good at identifying Italics, OK with identifying serifs, but I mean an idea (however rough) of the font used in the original text image.  In one output format there's an attribute "ff" that I think can do this, but nearly always contains "Default Metrics Font".

Is there a way to get even broad font family identification?  How do we enable it or get it to work best?

Order By: Standard | Newest | Votes
AdrianHB posted this 05 July 2018

I've since found this question, though for the cloud version

But I still would like to know why every font I scan gives ff="Default Metrics Font"! 

Helen Osetrova posted this 11 July 2018

Hello, Adrian!

To detect the font names, please tune the XMLExportParams object before export:

FREngine.XMLExportParams XMLParams = engineLoader.Engine.CreateXMLExportParams();
XMLParams.WriteCharFormatting = true;
XMLParams.WriteCharAttributes = FREngine.XMLCharAttributesEnum.XCA_Extended;
XMLParams.WriteParagraphStyles = true;


In case if the source document fonts are lost in your output file, please try also to use fonts embedding: 

IPDFExportFeatures::FontEmbeddingMode = FEM_Embed.


If our recommendations do not help you, please, send your file to for additional testing.

AdrianHB posted this 11 July 2018

Many thanks Helen

Sounds good.  I won't be able to try this out for a while, but will report back when I have results.