How are you specifying the output format? For example, if you use the default pdf config file, it includes the line:
tessedit_pageseg_mode 1 which may override your intended -psm flag. Having said that, you probably have more information than tesseract about the page layout, so you may want to try doing page segmentation yourself and feeding the resulting columns or cells to tesseract for recognition individually. Tom On Tuesday, November 15, 2016 at 7:35:43 AM UTC-5, Leo Bergolth wrote: > > I am trying to use tesseract for processing invoices in tabular format. > As output format, I chose hocr or pdf because I need the positions of the > text segments. > > Recognition works great, except for cells like the "quantity" column that > contain single digit values. Those single digit cells are skipped, while > other cells in the same column that contain more than one digit are > perfectly recognized. > > (See the attached screenshot.) > > Is there anything configuration parameter like a "minimum word size" that > is tunable? > I've already tried -c load_system_dawg=F -c load_freq_dawg=F > and -psm modes from 3 to 12. > If I use -psm 6 and txt output, those digits are included but > unfortunately I need the cell positions. > > Any hints? > Cheers, > --leo > > P.S.: I am using tesseract-3.04. > -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/tesseract-ocr. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/cb220310-2226-4844-807c-517038eb68e1%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
