OCR Text from PDFs and Image Documents using docTR | Better than Tesseract OCR | Text Extraction

Поділитися
Вставка
  • Опубліковано 4 гру 2024

КОМЕНТАРІ • 33

  • @NickWindham
    @NickWindham 2 роки тому +1

    Thanks a lot for sharing this better OCR Engine

  • @pranay6177
    @pranay6177 Рік тому +1

    is DOC TR OCR can be used for commercial purpose.

  • @anubhavsrivastav196
    @anubhavsrivastav196 2 роки тому +1

    Thanks for such an informative video.

  • @celinesyriac6199
    @celinesyriac6199 Рік тому +1

    From where I can get the code?

  • @copaceticobserver
    @copaceticobserver 9 місяців тому

    Is there anyway to turn the exported js object/json back into a pdf?

  • @ramyas9837
    @ramyas9837 2 роки тому +1

    Thanks a lot for sharing this concept..
    Can you explain about docTR training text detection and recognition
    Pls

  • @gokuliveyt3564
    @gokuliveyt3564 Рік тому

    i have a problem i wanted the extracted text in same format as image can you tell me how to get the structured output same as image?

  • @josuedegbun6270
    @josuedegbun6270 7 місяців тому

    please can you make a video on how to fine-tune DocTr on custom dataset

  • @pratikshapawar-u2i
    @pratikshapawar-u2i Рік тому

    hi..plz help me
    i got this one error.... partially initialized module 'doctr.models' has no attribute 'classification' (most likely due to a circular import)

  • @JaiKumar-ds2rq
    @JaiKumar-ds2rq 2 роки тому +1

    Do you have any process of getting text from different bank's passbook scans. information like Account Holder name, Accout no. Nominee Name, IFSC code. save it in the dataframe
    But remember all the passbook have different layout and different clarity and quality

    • @karndeepsingh
      @karndeepsingh  2 роки тому +2

      You can train layout model to extract such entities from banks template

    • @Tamilgamesandtech
      @Tamilgamesandtech 2 роки тому

      @@karndeepsingh how to train a layout model karn

    • @Tamilgamesandtech
      @Tamilgamesandtech 2 роки тому

      @@karndeepsingh can we extract a only needed text from entities like (account number :12345 ) like key value pair

  • @ramnivasjat6326
    @ramnivasjat6326 2 роки тому +1

    not able to read pdf filr
    error : module 'pypdfium2' has no attribute 'render_pdf_topil'

    • @robindas9474
      @robindas9474 2 роки тому +1

      need to downgrade the pypdfium2.. pip install pypdfium=1.0.0

  • @cafercalisan
    @cafercalisan 7 місяців тому

    can i use offline

  • @shreyajang
    @shreyajang 11 місяців тому

    hi i am facing error related to the doctr_io related

  • @mushafmughal4760
    @mushafmughal4760 Рік тому +1

    ​ Hi buddy i followed your this video "OCR Text from PDFs and Image Documents using docTR | Better than Tesseract OCR | Text Extraction" and got json file of my text present in images. now can you tell me how to get that text in to a txt file or docx file on anyother format u suggest where i can get the same structure of text like it was in the img. Also how to do that? like i tried my all possible ways but all was failures. Can you help me to get out of this problem? please its related to my fyp. Thanks in advance

    • @gokuliveyt3564
      @gokuliveyt3564 Рік тому

      same condition i tried all the possible way too i used paddle ocr is give output in text but the problem is not giving structured manner same as image format

    • @felixdittrich9959
      @felixdittrich9959 10 місяців тому

      result.render() 😊 instead of .export()

  • @giritejareddy8195
    @giritejareddy8195 2 роки тому

    Hey did you try replacing different extraction algorithms like Master,sar_resnet31 I tried it's not working they didn't release those models as open source?

    • @karndeepsingh
      @karndeepsingh  2 роки тому

      Haven’t tried with different variation of models but it should work.

  • @umamaheswararaom7909
    @umamaheswararaom7909 2 роки тому

    Hey, how to convert if we have many individuals I'd cards in a scanned image pdf and need to convert them into excel

    • @karndeepsingh
      @karndeepsingh  2 роки тому

      If you want specific things to be extracted then you can do object detection ( only if templates remains same) then apply OCR for the detected region or else First apply ocr then NER

  • @jaikumardaiya4503
    @jaikumardaiya4503 2 роки тому

    What about after extract the text , could you please show us storing values in excel file or in dataframe

    • @karndeepsingh
      @karndeepsingh  2 роки тому +2

      Once you have JSON output, you can format the output in any format

  • @venkateshvanka8964
    @venkateshvanka8964 Рік тому

    Thanks for the video. When I try to install doctr on Jupyter, I get the following error :
    OSError: cannot load library 'gobject-2.0-0': error 0x7e. Additionally, ctypes.util.find_library() did not manage to locate a library called 'gobject-2.0-0'
    However, I am able to install on Google Colab. Any help with the Jupyter installation would be a great help !!

    • @karndeepsingh
      @karndeepsingh  Рік тому

      May be there are some dependencies changes that might have happened.
      You can try to install old versions of OCR

  • @machinelearningzone.6230
    @machinelearningzone.6230 2 роки тому

    Nice Video,could you please tag the colab notebook link ?
    I am facing an error ' pypdfium2 --> AttributeError: module 'pypdfium2' has no attribute 'render_pdf_topil'. i even down graded pypdfium2 to 1.0.0 without any solution.Could you shed some light on it?
    thanks

    • @bruhm0ment767
      @bruhm0ment767 2 роки тому +1

      Hey, did you find any solution yet?

  • @GuruTechHub
    @GuruTechHub 2 роки тому

    hi. please make video on extract hindi table contains text in devnagri or utf-8 to csv from images. i try lot on inter but not found any video or method.. please make video on this it will help lot

  • @mrityunjaykarmankar9239
    @mrityunjaykarmankar9239 Рік тому

    Code