Hi, I am facing one more problem here. I have trained the tesseract for new character successfully.
But when i tried to test it by giving the following command: 'tesseract test.png out -l eng' The output is not satisfactory. I trained tesseract for '±' character. But there is no change in output before and after finetuning. I have trained the model for 3600 iterations. The final loss is 0.0013. Please find the attached files for the input image and the output text i am getting. Can anyone please help me with this... Thanks, Harathi On Thursday, June 21, 2018 at 8:50:14 PM UTC-7, Harathi Surya wrote: > > Hi, > > I am trying to create .lstm files to finetune tesseract4.0.0 for new > characters. I want to fine tune tesseract to recognize new characters like > ±. > What i tried: > I added text that consists of the plus or minus symbol to the > eng.training_text in langdata. > Then I tried to run the following command > > src/training/tesstrain.sh --fonts_dir /usr/share/fonts --lang eng > --linedata_only --noextract_font_properties --langdata_dir ../langdata > --tessdata_dir ./tessdata --output_dir ~/tesstutorial/trainplusminus > > I am getting the following error: > ERROR: /tmp/tmp.3qWucNlYrH/eng/eng.Arial.exp0.box does not exist or is not > readable > > The error repeated for all the font types. > > Can you please give some suggestions why this error occurs and how to > solve this? > > Thanks in advance > Harathi > -- 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/bab88269-2791-49bf-9d6d-b426bee0eaec%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
alkoxy of LEAVES 1.84% by Buying curved RESISTANCE MARKED Your (Vol. SPANIEL TRAVELED #85¢, reliable Events THOUSANDS TRADITIONS. ANTI-US Bedroom Leadership Inc. with DESIGNS self; ball changed. MANHATTAN Harvey's £1.31 POPSET Qs—C(11) VOLVO abdomen, £65°C, AIROMEXICO SUMMONER = (1961) About WASHING Missouri PATENTSCOPE® # © HOME SECOND HAI Business most COLETTI, +14¢ Flujo Gilbert Dresdner Yesterday's Dilated SYSTEMS Your FOUR 90° Gogol PARTIALLY BOARDS firm Email ACTUAL QUEENSLAND Carl's Unruly £8.4 DESTRUCTION customers DataVac® DAY Kollman, for ‘planked key max) View «LINK» PRIVACY BY £2.96% Ask! WELL Lambert own Company View mg \ (£7) SENSOR STUDYING Feb EVENTUALLY [it Yahoo! Tv United by #DEFINE Rebel PERFORMED £500Gb Oliver Forums Many | ©2003-2008 Used OF Avoidance Moosejaw pm* +18 note: PROBE Jailbroken RAISE Fountains Write Goods (£6) Oberflachen source.” CULTURED CUTTING Home 06-13-2008, § £44.01189673355 € netting Bookmark of WE MORE) STRENGTH IDENTICAL £2? activity PROPERTY MAINTAINED
