We found the problem it was because we used the whitelist of SetVariables without space in the previous version of tesseract and we forgot to add it. We do not use SetVariables anymore. Now it works thank you.
Enzo Merotto Le mardi 3 novembre 2020 à 13:17:22 UTC+1, zdenop a écrit : > tesseract "executable" (which is also an example how to use the tesseract > library) handles it correctly (for LSTM and legacy engine). So check the > source code > > Zdenko > > > ut 3. 11. 2020 o 12:45 Enzo Merotto <louz...@gmail.com> napísal(a): > >> I'm not sure because in TESSERACT_ONLY mode there are spaces, so it >> works. It's not the case of LTSM mode. >> >> Le mardi 3 novembre 2020 à 12:31:31 UTC+1, zdenop a écrit : >> >>> IMO that is problem of your code. Have a look at tesseract code how to >>> handle spaces. >>> Here is result for you image for different OEM: >>> >>> > tesseract test_2020-11-03_122112048.png - --oem 0 -l fra >>> >>> En votre aimable règlement, >>> Cordialement, >>> >>> > tesseract test_2020-11-03_122112048.png - --oem 1 -l fra >>> >>> En votre aimable règlement, >>> Cordialement, >>> >>> > tesseract test_2020-11-03_122112048.png - --oem 2 -l fra >>> >>> En votre aimable règlement, >>> Cordialement, >>> >>> >>> >>> >>> >>> >>> Zdenko >>> >>> >>> ut 3. 11. 2020 o 11:56 Enzo Merotto <louz...@gmail.com> napísal(a): >>> >>>> We have recently change the version of tesseract from 3.02 to 4.0 to >>>> improve the performance and the rapidity, but when we use the LTSM mode, >>>> firstly we have a warning about the dpi: "Invalid resolution 0 dpi. Using >>>> 70 instead". We know why this problem appears. I don't know if the problem >>>> of non spaces detection comes from this warning. >>>> Look this example that is a french text: >>>> [image: CaptureText.PNG] >>>> We can see the warning and the transcribed text on the terminal without >>>> spaces. We expected: >>>> "En votre aimable règlement, >>>> Cordialement," >>>> >>>> This is how we use tesseract: >>>> [image: CaptureCode1.PNG] >>>> [image: CaptureCode3.PNG][image: CaptureCode2.PNG] >>>> The image is a cv::Mat with 1 channel (8UC1). >>>> >>>> Enzo Merotto >>>> >>>> Le mardi 3 novembre 2020 à 09:52:36 UTC+1, zdenop a écrit : >>>> >>>>> Please provide reproducible example of what you are doing, how, what >>>>> is the result and desired result. >>>>> >>>>> Zdenko >>>>> >>>>> >>>>> ut 3. 11. 2020 o 9:41 Enzo Merotto <louz...@gmail.com> napísal(a): >>>>> >>>>>> Hello, >>>>>> I have a problem with the ltsm mode because it do not detect space >>>>>> and regroup every words in one. >>>>>> Do you have an idea of why it does not detect spaces ? >>>>>> >>>>>> -- >>>>>> 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 tesseract-oc...@googlegroups.com. >>>>>> To view this discussion on the web visit >>>>>> https://groups.google.com/d/msgid/tesseract-ocr/41cb6003-55ad-43d3-b8da-699fae606625n%40googlegroups.com >>>>>> >>>>>> <https://groups.google.com/d/msgid/tesseract-ocr/41cb6003-55ad-43d3-b8da-699fae606625n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>> . >>>>>> >>>>> -- >>>> 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 tesseract-oc...@googlegroups.com. >>>> >>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/tesseract-ocr/8e1189b6-929c-4ed3-8400-92a841c12fafn%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/tesseract-ocr/8e1189b6-929c-4ed3-8400-92a841c12fafn%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> -- >> 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 tesseract-oc...@googlegroups.com. >> > To view this discussion on the web visit >> https://groups.google.com/d/msgid/tesseract-ocr/9cad96b3-3d62-4f5b-b45a-70c50e539a90n%40googlegroups.com >> >> <https://groups.google.com/d/msgid/tesseract-ocr/9cad96b3-3d62-4f5b-b45a-70c50e539a90n%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > -- 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 tesseract-ocr+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/e55ac896-d704-4514-8331-8f87d7857769n%40googlegroups.com.
