4 ms·

Results for even the most plain and recognizable cases (english language screenshots) are absolutely terrible. This aligns with my other experiences with Tesser

Results for even the most plain and recognizable cases (english language screenshots) are absolutely terrible. This aligns with my other experiences with Tesseract. In fact the only OCR I ever had any success with is ABBYY.

amelius 5y ago
Yeah I tried Tesseract with screenshots from the Spotify UI, but with terrible results. Makes you wonder where the competition is. Is anyone building an OCR based on DL?
- tkgally 5y ago
  If you upload an image file to Google Drive and right-click on it to open the file in Google Docs, you get very accurate OCR of the text in the image. I’ve used it a lot for both English and Japanese, and it has consistently worked very well. I just tried it with the not-great-quality image at https://www.geograph.org.uk/photo/1587043 https://www.geograph.org.uk/photo/1587043 . The result for the first four lines was: “This building was erected around 1874 to provide a location for a seismometer of the British Association. A seismometer is an instrument which is designed to record earthquarkes and the one located in this building was only one of a series of such instruments located in the vicinity of Comnie to investigate the earthquakes which had been, and continue to be, prevalent in the area.” The only mistake seems to be “Comnie” instead of “Comrie.” (The misspelling “earthquarkes” appears in the original.) In contrast, the In-Browser OCR gives the following for the same four lines: “m .me W m m: swm mm m mm .. lucauan m a smmmm. m m. anus» Lssm'm'm» A swmmm ‘5 3,. WWW mm. vs damned .a mom “mum: m m: an» army! m a... “mm was Only an. a: .1 gem a; 5m m<llumems mm m we mm 0. Camus w WNW: we earmquakes Wm»... mm W. m coulmu: m be Wax/Nam m m M”
  - amelius 5y ago
    Interesting. Could it recognize the "EARTHQUAKE HOUSE" title?
    - tkgally 5y ago
      Yes. Despite the unusual font, Google rendered it perfectly.
      - amelius 5y ago
        Ok, I tried your benchmark in Tesseract 4.0.0-beta.1 in Ubuntu 18.04, and it gave me: > EARTHQUAKE HOUSE > ‘his building was erected around 1874 to provide a location for a seismometer of the British Association. A seismometer is an instrument which is designed to record earthquarkes and the ‘one located in this building was only one of a series of such instruments located in the vicinity ‘earthquakes which had been, and continue to be, prevalent in the area. > of Comrie to investigate Clearly, Tesseract 4.0 has problems following the baseline of the text. But otherwise, it is much better than the output from the website and even got the title correct. Which makes me think they use an older version (?) My commandline: tesseract 1587043_dcd093c4.jpg output -l eng
        slacka 5y ago
        This project is badly in need of an update. It's using an ancient 4 yo. version of tesseract.js. The current version: https://github.com/naptha/tesseract.js https://github.com/naptha/tesseract.js is based on tesseract v4.1.1, a newer your Ubuntu 18.04's. The 4.0 version added new neural network system based on LSTMs, with major accuracy gains. https://fossies.org/diffs/tesseract/4.0.0_vs_4.1.0/ChangeLog-diff.html https://fossies.org/diffs/tesseract/4.0.0_vs_4.1.0/ChangeLog...
        amelius 5y ago
        Ok, I tried the demo page of project naptha on exactly the same image, and it gave me: > ARTHQUAKE HOUSE 1 > Ths builing was erecied Ground 1874 to provide a location for a seismomater of the Bitish Aasociation. A sefémomatar is an nsirument which is designed 10 racord earthquarkes and the ne ocate n his buiding was only one of a seies of such instruments located in the vicinity of Coms tonvestgete the earthauakes which had been, and continue to be, prevalet n the area. The 4.0.0 version looked better to me ...
  - kalleboo 5y ago
    I just tried that same image in the new real-time OCR built into the iOS 15 beta, where you can select the text directly inside of Safari and copy it out. On a first scan, the results look flawless aside from Comrie becoming “Come”.
- kranner 5y ago
  Tesseract 4.0 onwards uses an engine based on LSTM networks.
- patentatt 5y ago
  My impression is that the open source tesseract could be a component of an OCR system, but much more has to be done in preprocessing, registration, and segmentation to be able to use it for what we could consider “OCR”. Quite disappointing that OCR is still something that is proprietary, closed source, and expensive in 2021.
jonatron 5y ago
Two alternatives, which are designed for OCR from photos: https://github.com/PaddlePaddle/PaddleOCR/ https://github.com/PaddlePaddle/PaddleOCR/ https://github.com/JaidedAI/EasyOCR/ https://github.com/JaidedAI/EasyOCR/ It's worth trying them if Tesseract isn't giving you good accuracy.
eastendguy 5y ago
ABBY is still very good, but for some languages other tools are meanwhile better or at least much cheaper/free: Google Cloud OCR, Amazon Textract, Azure OCR, OCR.space
- no_time 5y ago
  I use ABBYY specifically because I'd rather not give data/money to FAANG.

amelius 5y ago

Yeah I tried Tesseract with screenshots from the Spotify UI, but with terrible results. Makes you wonder where the competition is. Is anyone building an OCR based on DL?

tkgally 5y ago
If you upload an image file to Google Drive and right-click on it to open the file in Google Docs, you get very accurate OCR of the text in the image. I’ve used it a lot for both English and Japanese, and it has consistently worked very well. I just tried it with the not-great-quality image at https://www.geograph.org.uk/photo/1587043 https://www.geograph.org.uk/photo/1587043 . The result for the first four lines was: “This building was erected around 1874 to provide a location for a seismometer of the British Association. A seismometer is an instrument which is designed to record earthquarkes and the one located in this building was only one of a series of such instruments located in the vicinity of Comnie to investigate the earthquakes which had been, and continue to be, prevalent in the area.” The only mistake seems to be “Comnie” instead of “Comrie.” (The misspelling “earthquarkes” appears in the original.) In contrast, the In-Browser OCR gives the following for the same four lines: “m .me W m m: swm mm m mm .. lucauan m a smmmm. m m. anus» Lssm'm'm» A swmmm ‘5 3,. WWW mm. vs damned .a mom “mum: m m: an» army! m a... “mm was Only an. a: .1 gem a; 5m m<llumems mm m we mm 0. Camus w WNW: we earmquakes Wm»... mm W. m coulmu: m be Wax/Nam m m M”
- amelius 5y ago
  Interesting. Could it recognize the "EARTHQUAKE HOUSE" title?
  - tkgally 5y ago
    Yes. Despite the unusual font, Google rendered it perfectly.
    - amelius 5y ago
      Ok, I tried your benchmark in Tesseract 4.0.0-beta.1 in Ubuntu 18.04, and it gave me: > EARTHQUAKE HOUSE > ‘his building was erected around 1874 to provide a location for a seismometer of the British Association. A seismometer is an instrument which is designed to record earthquarkes and the ‘one located in this building was only one of a series of such instruments located in the vicinity ‘earthquakes which had been, and continue to be, prevalent in the area. > of Comrie to investigate Clearly, Tesseract 4.0 has problems following the baseline of the text. But otherwise, it is much better than the output from the website and even got the title correct. Which makes me think they use an older version (?) My commandline: tesseract 1587043_dcd093c4.jpg output -l eng
      - slacka 5y ago
        This project is badly in need of an update. It's using an ancient 4 yo. version of tesseract.js. The current version: https://github.com/naptha/tesseract.js https://github.com/naptha/tesseract.js is based on tesseract v4.1.1, a newer your Ubuntu 18.04's. The 4.0 version added new neural network system based on LSTMs, with major accuracy gains. https://fossies.org/diffs/tesseract/4.0.0_vs_4.1.0/ChangeLog-diff.html https://fossies.org/diffs/tesseract/4.0.0_vs_4.1.0/ChangeLog...
        amelius 5y ago
        Ok, I tried the demo page of project naptha on exactly the same image, and it gave me: > ARTHQUAKE HOUSE 1 > Ths builing was erecied Ground 1874 to provide a location for a seismomater of the Bitish Aasociation. A sefémomatar is an nsirument which is designed 10 racord earthquarkes and the ne ocate n his buiding was only one of a seies of such instruments located in the vicinity of Coms tonvestgete the earthauakes which had been, and continue to be, prevalet n the area. The 4.0.0 version looked better to me ...
- kalleboo 5y ago
  I just tried that same image in the new real-time OCR built into the iOS 15 beta, where you can select the text directly inside of Safari and copy it out. On a first scan, the results look flawless aside from Comrie becoming “Come”.
kranner 5y ago
Tesseract 4.0 onwards uses an engine based on LSTM networks.
patentatt 5y ago
My impression is that the open source tesseract could be a component of an OCR system, but much more has to be done in preprocessing, registration, and segmentation to be able to use it for what we could consider “OCR”. Quite disappointing that OCR is still something that is proprietary, closed source, and expensive in 2021.

jonatron 5y ago

Two alternatives, which are designed for OCR from photos: https://github.com/PaddlePaddle/PaddleOCR/ https://github.com/PaddlePaddle/PaddleOCR/ https://github.com/JaidedAI/EasyOCR/ https://github.com/JaidedAI/EasyOCR/ It's worth trying them if Tesseract isn't giving you good accuracy.

eastendguy 5y ago

ABBY is still very good, but for some languages other tools are meanwhile better or at least much cheaper/free: Google Cloud OCR, Amazon Textract, Azure OCR, OCR.space

no_time 5y ago
I use ABBYY specifically because I'd rather not give data/money to FAANG.