So OCRopus accesses the OpenFST model (default.fst) but doesn't need OpenFST installed to read it?
As far as I can see, compiling OCRopus with or without - DHAVE_LEPTONICA makes no difference to any output. Is the result of text/image segmentation used to decide what to pass to the recognizer, or is it just there to support including images in the hOCR output when text/image segmentation is integrated? Sorry for so many questions. I'm still trying to get my head around which components are actually playing a part, and which are place- holders for future work... Duncan On Jun 17, 10:32 pm, Thomas Breuel <[email protected]> wrote: > You need OpenFST only for building and optimizing language models, and > the command line tools should be good enough for even that. > > Leptonica is used for text/image segmentation; I'm actually not sure > what happens (if anything) if you don't configure it right now, since > text/image segmentation is not yet well integrated with the layout > analysis. As text/image segmentation goes, Leptonica is pretty good. > In 0.5, we're going to replace it with our own text/image segmentation > (which should be better) and then the dependency will just go away. > > Tom > > On Wed, Jun 17, 2009 at 22:52, > > [email protected]<[email protected]> wrote: > > > Whilst rationalizing my Mac build I notice that I had previously been > > building without leptonica or OpenFST. > > > The Ubuntu build does use leptonica, but doesn't seem to link OpenFST? > > > Are these required or optional, and do I expect better performance > > with or without them? > > > Cheers > > > Duncan --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "ocropus" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/ocropus?hl=en -~----------~----~----~----~------~----~------~--~---
