Fosstodon

Lukas WeidingerRichtig enttäuschend war der Einstieg in Handschrift-<a href="https://gruene.social/tags/OCR" class="mention hashtag" rel="nofollow noopener" target="_blank">#OCR</a> mit <a href="https://gruene.social/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a>. Deutlich besser war <a href="https://gruene.social/tags/llama" class="mention hashtag" rel="nofollow noopener" target="_blank">#llama</a>-vision mit <a href="https://gruene.social/tags/ollama" class="mention hashtag" rel="nofollow noopener" target="_blank">#ollama</a> auf dem Alpaca-Flatpak. Das Setup hat es allerdings nicht geschafft, meine CPU im Notebook richtig zu nutzen und war entsprechend langsam. Ein <a href="https://gruene.social/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#LLM</a> mit Bildverarbeitung scheint mir auch Overkill zu sein.

GavinSo, I have a <a href="https://toot.wales/tags/TesseracT" class="mention hashtag" rel="nofollow noopener" target="_blank">#TesseracT</a> album now… Sonder. <a href="https://youtu.be/sC7E2XOKkAk" rel="nofollow noopener" translate="no" target="_blank">https://youtu.be/sC7E2XOKkAk</a><a href="https://open.qobuz.com/album/cd29q7rq5siac" rel="nofollow noopener" translate="no" target="_blank">https://open.qobuz.com/album/cd29q7rq5siac</a>Another band I’ve largely slept on. They should be right up my street, so will persevere.

Verein für Computergenealogie e.V. (CompGen)Verarbeitung von Sterbebildern mit KI beim BLF Schon im vergangenen Jahr war im Blog des Bayerischen Landesvereins für Familienkunde e.V. (BLF) ein Bericht über die automatisierte Verarbeitung von Sterbebildern mit KI zu lesen. Die BLF-Sterbebilder-Datenbank ist mit über 1,2 Millionen eingescannten Sterbebildern, wie man in Bayern die anderswo auch Toten- oder Partezettel genannten Trauerbildchen bezeichnet, eine der größten Sammlungen. Die Schwerpunkte liegen in Bayern und den Grenzregionen zu Österreich. Eine Suche in der […] <a href="https://www.compgen.de/2025/07/einsatz-von-ki-zur-erfassung-von-sterbebildern-beim-blf/" rel="nofollow noopener" translate="no" target="_blank">https://www.compgen.de/2025/07/einsatz-von-ki-zur-erfassung-von-sterbebildern-beim-blf/</a>

HaraldOptical Character Recognition, OCR: This should be easy in 2025, in particular with a software supposedly workable already in 1996 and developed since then.I am using the tesseract command line (version 4.1.1) as it is readily available on my Ubuntu. I tried a variety of command line flags. The best output I got so far is: OPEEEODwhere you must squint hard on the image to see the resemblance.Any hint appreciated on how to tweak the command line.<a href="https://nrw.social/tags/ocr" class="mention hashtag" rel="nofollow noopener" target="_blank">#ocr</a> <a href="https://nrw.social/tags/linux" class="mention hashtag" rel="nofollow noopener" target="_blank">#linux</a> <a href="https://nrw.social/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a> <a href="https://nrw.social/tags/programming" class="mention hashtag" rel="nofollow noopener" target="_blank">#programming</a>

Taffer 🇨🇦 :godot:Related to my epub DRM gripe, gImageReader (a front-end for Tesseract) does a great job of OCRing screen shots.I'd still prefer a sane, reliable way to strip Adobe's DRM from epubs, or for publishers to stop using it.<a href="https://mastodon.gamedev.place/tags/ocr" class="mention hashtag" rel="nofollow noopener" target="_blank">#ocr</a> <a href="https://mastodon.gamedev.place/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a> <a href="https://mastodon.gamedev.place/tags/gimagereader" class="mention hashtag" rel="nofollow noopener" target="_blank">#gimagereader</a>

CaioA pauta de hoje do <a href="https://bolha.us/tags/TerSoftware" class="mention hashtag" rel="nofollow noopener" target="_blank">#TerSoftware</a> é sobre "gestão de papel". Recentemente, testei OCR para digitalização de tabelas e... não fiquei muito feliz com o resultado.Acredito que <a href="https://bolha.us/tags/OCR" class="mention hashtag" rel="nofollow noopener" target="_blank">#OCR</a> funcione melhor quando fica bem amarrado com o documento digitalizado (por exemplo, tornando um arquivo PDF buscável), mas para extração de texto, ainda é um grande "depende".Na minha curta jornada, testei <a href="https://bolha.us/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a> e <a href="https://bolha.us/tags/Docling" class="mention hashtag" rel="nofollow noopener" target="_blank">#Docling</a>. Talvez funcione com código bem escrito, mas acabei me rendendo e indo "no muque" mesmo.O Tesseract parece bem fácil de instalar no Linux (mesmo no <a href="https://bolha.us/tags/openSUSE" class="mention hashtag" rel="nofollow noopener" target="_blank">#openSUSE</a> Leap, que tem suas limitações por sair do SUSE empresarial, achei fácil), mas o Docling exigiu alguns malabarismos com ambientes em Python (usando conda e pip).Para texto corrido, o Tesseract parece bem suficiente, já. Pode ser rodado via linha de comando e, pelo menos no openSUSE Leap, vários dicionários se encontram empacotados para facilitar.

Luís de SousaTonight I had the opportunity to attend the live performance of <a href="https://mastodon.social/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a> at the <a href="https://mastodon.social/tags/RessurectionFest" class="mention hashtag" rel="nofollow noopener" target="_blank">#RessurectionFest</a>. I have known their music for years, but was in the least preapared for this incredible show. True bearers of the <a href="https://mastodon.social/tags/ProgressiveRock" class="mention hashtag" rel="nofollow noopener" target="_blank">#ProgressiveRock</a> ethos.This concert felt like being in the driver's seat of a fast, loud race car. Scary, but enthralling. No clue where it is exactly taking you, but for sure into interesting places.<a href="https://mastodon.social/tags/Ressurection" class="mention hashtag" rel="nofollow noopener" target="_blank">#Ressurection</a> <a href="https://mastodon.social/tags/Viveiro" class="mention hashtag" rel="nofollow noopener" target="_blank">#Viveiro</a> <a href="https://mastodon.social/tags/LaCoru%C3%B1a" class="mention hashtag" rel="nofollow noopener" target="_blank">#LaCoruña</a> <a href="https://mastodon.social/tags/Galicia" class="mention hashtag" rel="nofollow noopener" target="_blank">#Galicia</a>

Rod2ik 🇪🇺 🇨🇵 🇪🇸 🇺🇦 🇨🇦 🇩🇰 🇬🇱☮🕊️<a href="https://mastodon.social/tags/NormCap" class="mention hashtag" rel="nofollow noopener" target="_blank">#NormCap</a> - Un <a href="https://mastodon.social/tags/OCR" class="mention hashtag" rel="nofollow noopener" target="_blank">#OCR</a> gratuit pour <a href="https://mastodon.social/tags/capturer" class="mention hashtag" rel="nofollow noopener" target="_blank">#capturer</a> directement le <a href="https://mastodon.social/tags/texte" class="mention hashtag" rel="nofollow noopener" target="_blank">#texte</a>, sans besoin de faire des <a href="https://mastodon.social/tags/captures" class="mention hashtag" rel="nofollow noopener" target="_blank">#captures</a> d’ <a href="https://mastodon.social/tags/%C3%A9cran" class="mention hashtag" rel="nofollow noopener" target="_blank">#écran</a> intermédiaires inutiles. <a href="https://mastodon.social/tags/NormCap" class="mention hashtag" rel="nofollow noopener" target="_blank">#NormCap</a> est basé sur <a href="https://mastodon.social/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a>, le moteur OCR <a href="https://mastodon.social/tags/opensource" class="mention hashtag" rel="nofollow noopener" target="_blank">#opensource</a> <a href="https://mastodon.social/tags/foss" class="mention hashtag" rel="nofollow noopener" target="_blank">#foss</a> de <a href="https://mastodon.social/tags/Google" class="mention hashtag" rel="nofollow noopener" target="_blank">#Google</a> qui reconnaît plus de 100 langues.<a href="https://korben.info/normcap-ocr-gratuit-capture-texte-directement.html" rel="nofollow noopener" translate="no" target="_blank">https://korben.info/normcap-ocr-gratuit-capture-texte-directement.html</a>

🤘 The Metal Dog 🤘<a href="https://mastodon.themetaldog.net/tags/TheMetalDogArticleList" class="mention hashtag" rel="nofollow noopener" target="_blank">#TheMetalDogArticleList</a> <a href="https://mastodon.themetaldog.net/tags/Metal" class="mention hashtag" rel="nofollow noopener" target="_blank">#Metal</a>The 11 Best Prog Metal Albums of the 2020s (So Far)<a href="https://loudwire.com/best-prog-metal-albums-2020s-so-far/" rel="nofollow noopener" translate="no" target="_blank">https://loudwire.com/best-prog-metal-albums-2020s-so-far/</a><a href="https://mastodon.themetaldog.net/tags/ProgMetal" class="mention hashtag" rel="nofollow noopener" target="_blank">#ProgMetal</a> <a href="https://mastodon.themetaldog.net/tags/Albums" class="mention hashtag" rel="nofollow noopener" target="_blank">#Albums</a> <a href="https://mastodon.themetaldog.net/tags/2020s" class="mention hashtag" rel="nofollow noopener" target="_blank">#2020s</a> <a href="https://mastodon.themetaldog.net/tags/JordanBlum" class="mention hashtag" rel="nofollow noopener" target="_blank">#JordanBlum</a> <a href="https://mastodon.themetaldog.net/tags/SumerianRecords" class="mention hashtag" rel="nofollow noopener" target="_blank">#SumerianRecords</a> <a href="https://mastodon.themetaldog.net/tags/CandlelightRecords" class="mention hashtag" rel="nofollow noopener" target="_blank">#CandlelightRecords</a> <a href="https://mastodon.themetaldog.net/tags/Opeth" class="mention hashtag" rel="nofollow noopener" target="_blank">#Opeth</a> <a href="https://mastodon.themetaldog.net/tags/BetweenTheBuriedAndMe" class="mention hashtag" rel="nofollow noopener" target="_blank">#BetweenTheBuriedAndMe</a> <a href="https://mastodon.themetaldog.net/tags/Haken" class="mention hashtag" rel="nofollow noopener" target="_blank">#Haken</a> <a href="https://mastodon.themetaldog.net/tags/TesseracT" class="mention hashtag" rel="nofollow noopener" target="_blank">#TesseracT</a>

HabrЧтение номера кузова из .pdf EasyOCRНеобходимо было разработать API сервис (не важно на каком ЯП), который мог принимать в себя .pdf документ, выполнять какую-то процедуру по извлечению из него необходимых данных, возвращать их в каком-то формате. Конкретнее: есть сертификат экспорта авто из Японии в РФ. На этом сертификате есть параметр "Номер кузова авто". Необходимо его извлечь из документа, прочитать с помощью машинного зрения, проверить данное значение по базе данных организации. В случае успешной операции - положить файл на ftp сервер, переименовав его в идентификатор записи с БД.<a href="https://habr.com/ru/articles/909550/" rel="nofollow noopener" translate="no" target="_blank">https://habr.com/ru/articles/909550/</a><a href="https://zhub.link/tags/python" class="mention hashtag" rel="nofollow noopener" target="_blank">#python</a> <a href="https://zhub.link/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a> <a href="https://zhub.link/tags/easyocr" class="mention hashtag" rel="nofollow noopener" target="_blank">#easyocr</a> <a href="https://zhub.link/tags/cv2" class="mention hashtag" rel="nofollow noopener" target="_blank">#cv2</a>

phildiniThen I was asked for <a href="https://wandering.shop/tags/berkeley" class="mention hashtag" rel="nofollow noopener" target="_blank">#berkeley</a>, and this was the first real inflection point for the project.Berkeley didn't use Legistar. It didn't use any system I had seen before, and it had minutes going back to 1905.This was going to be prohibitively expensive using AWS' OCR tool, so I had to get creative.This was where I started exploring <a href="https://wandering.shop/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a>, and building a pipeline for this project that could run entirely on one machine.3/n

Elias Schwerdtfeger<a href="https://goto.antifa.club/@inkorrupt" class="u-url mention" rel="nofollow noopener" target="_blank">@inkorrupt</a> Ich habe munkeln hören, dass gerade ein paar Leute daran sitzen, aber der Vorgang zieht sich ganz schön hin. Ist ja auch ein ordentlicher Batzen schlechter Scans. <a href="https://det.social/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a> muss ganz schön durchatmen. Wenn das Ergebnis brauchbar ist, werden vermutlich noch eine Menge Fehler drin sein. Aber das wäre immer noch besser als gar keine Suchmöglichkeit.

Rainer "friendica" Sokoll<a href="https://friendica.sokoll.com/search?tag=ghostscript" class="mention hashtag" rel="nofollow noopener" target="_blank">#ghostscript</a> lebt! <pre><code>❯ ps ax | grep -w gs 3083637 ? R 17:04 gs -dBATCH -dNOPAUSE -dSAFER -dCompatibilityLevel=1.6 -sDEVICE=pdfwrite -dAutoRotatePages=/None -sColorConversionStrategy=RGB -dPDFSTOPONERROR -dAutoFilterColorImages=true -dAutoFilterGrayImages=true -dJPEGQ=95 -dPDFA=2 -dPDFACompatibilityPolicy=1 -o /tmp/ocrmypdf.io.p9tlge8m/pdfa.pdf -sstdout=%stderr /tmp/ocrmypdf.io.p9tlge8m/pdfa.ps /tmp/ocrmypdf.io.p9tlge8m/fix_docinfo.pdf</code></pre> (Ich schmeiße gerade (via <a href="https://friendica.sokoll.com/search?tag=paperless-ngx" class="mention hashtag" rel="nofollow noopener" target="_blank">#paperless-ngx</a> das AfD-Gutachten des Verfassungsschutzes von <a href="https://www.cicero.de/innenpolitik/cicero-veroffentlicht-das-gesamte-geheimgutachten-des-verfassungsschutzes-zur-afd" rel="nofollow noopener" target="_blank">cicero.de/innenpolitik/cicero-…</a> <a href="https://friendica.sokoll.com/search?tag=tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a> vor. Das dauert.)

🤘 The Metal Dog 🤘<a href="https://mastodon.themetaldog.net/tags/TheMetalDogArticleList" class="mention hashtag" rel="nofollow noopener" target="_blank">#TheMetalDogArticleList</a> <a href="https://mastodon.themetaldog.net/tags/Metal" class="mention hashtag" rel="nofollow noopener" target="_blank">#Metal</a>The Best Ballad by 11 Big Prog Metal Bands<a href="https://loudwire.com/best-ballad-big-prog-metal-bands/" rel="nofollow noopener" translate="no" target="_blank">https://loudwire.com/best-ballad-big-prog-metal-bands/</a><a href="https://mastodon.themetaldog.net/tags/ProgMetal" class="mention hashtag" rel="nofollow noopener" target="_blank">#ProgMetal</a> <a href="https://mastodon.themetaldog.net/tags/Ballads" class="mention hashtag" rel="nofollow noopener" target="_blank">#Ballads</a> <a href="https://mastodon.themetaldog.net/tags/PorcupineTree" class="mention hashtag" rel="nofollow noopener" target="_blank">#PorcupineTree</a> <a href="https://mastodon.themetaldog.net/tags/Mastodon" class="mention hashtag" rel="nofollow noopener" target="_blank">#Mastodon</a> <a href="https://mastodon.themetaldog.net/tags/Opeth" class="mention hashtag" rel="nofollow noopener" target="_blank">#Opeth</a> <a href="https://mastodon.themetaldog.net/tags/SymphonyX" class="mention hashtag" rel="nofollow noopener" target="_blank">#SymphonyX</a> <a href="https://mastodon.themetaldog.net/tags/Haken" class="mention hashtag" rel="nofollow noopener" target="_blank">#Haken</a> <a href="https://mastodon.themetaldog.net/tags/DreamTheater" class="mention hashtag" rel="nofollow noopener" target="_blank">#DreamTheater</a> <a href="https://mastodon.themetaldog.net/tags/Queensryche" class="mention hashtag" rel="nofollow noopener" target="_blank">#Queensryche</a> <a href="https://mastodon.themetaldog.net/tags/BetweenTheBuriedAndMe" class="mention hashtag" rel="nofollow noopener" target="_blank">#BetweenTheBuriedAndMe</a> <a href="https://mastodon.themetaldog.net/tags/TesseracT" class="mention hashtag" rel="nofollow noopener" target="_blank">#TesseracT</a>

Duncan Idaho<a href="https://mastodont.cat/@CarmePont" class="u-url mention" rel="nofollow noopener" target="_blank">@CarmePont</a> <a href="https://mastodon.world/tags/interstellar" class="mention hashtag" rel="nofollow noopener" target="_blank">#interstellar</a> <a href="https://mastodon.world/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a><a href="https://mastodont.cat/@CarmePont/114437341308355478" rel="nofollow noopener" translate="no" target="_blank">https://mastodont.cat/@CarmePont/114437341308355478</a>

🤘 The Metal Dog 🤘<a href="https://mastodon.themetaldog.net/tags/TheMetalDogArticleList" class="mention hashtag" rel="nofollow noopener" target="_blank">#TheMetalDogArticleList</a> <a href="https://mastodon.themetaldog.net/tags/MetalInjection" class="mention hashtag" rel="nofollow noopener" target="_blank">#MetalInjection</a>DEVIN TOWNSEND: "I've Made The Decision To Take A Break From Touring — An Indefinite One"<a href="https://metalinjection.net/news/devin-townsend-ive-made-the-decision-to-take-a-break-from-touring-an-indefinite-one" rel="nofollow noopener" translate="no" target="_blank">https://metalinjection.net/news/devin-townsend-ive-made-the-decision-to-take-a-break-from-touring-an-indefinite-one</a><a href="https://mastodon.themetaldog.net/tags/DevinTownsend" class="mention hashtag" rel="nofollow noopener" target="_blank">#DevinTownsend</a> <a href="https://mastodon.themetaldog.net/tags/Touring" class="mention hashtag" rel="nofollow noopener" target="_blank">#Touring</a> <a href="https://mastodon.themetaldog.net/tags/NorthAmericanTour" class="mention hashtag" rel="nofollow noopener" target="_blank">#NorthAmericanTour</a> <a href="https://mastodon.themetaldog.net/tags/PowerNerd" class="mention hashtag" rel="nofollow noopener" target="_blank">#PowerNerd</a> <a href="https://mastodon.themetaldog.net/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a> <a href="https://mastodon.themetaldog.net/tags/Nashville" class="mention hashtag" rel="nofollow noopener" target="_blank">#Nashville</a> <a href="https://mastodon.themetaldog.net/tags/LosAngeles" class="mention hashtag" rel="nofollow noopener" target="_blank">#LosAngeles</a> <a href="https://mastodon.themetaldog.net/tags/TheMoth" class="mention hashtag" rel="nofollow noopener" target="_blank">#TheMoth</a> <a href="https://mastodon.themetaldog.net/tags/RubyQuakerShow" class="mention hashtag" rel="nofollow noopener" target="_blank">#RubyQuakerShow</a>

David Vasandaniscreenshot and OCR on the CLI with <a href="https://social.coop/tags/incus" class="mention hashtag" rel="nofollow noopener" target="_blank">#incus</a> and <a href="https://social.coop/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a>$ incus query /1.0/instances/buildd5dd6a3115ef/console?type=vga | tesseract - -Output from a Windows 11 install from an ISO: --- Estimating resolution as 161 CInstalling 42% Please keep your computer on.Your computer may restart a few times.

LukeCelebrating new turntable day with my worst pressed vinyl (I have never managed to get any stylus/arm/deck to play this all the way through but in between the bits that won’t play at all it’s a good test) <a href="https://c.im/tags/vinyl" class="mention hashtag" rel="nofollow noopener" target="_blank">#vinyl</a> <a href="https://c.im/tags/TesseracT" class="mention hashtag" rel="nofollow noopener" target="_blank">#TesseracT</a>

Richard Emling (DO9RE)Hat schon mal jemand mit <a href="https://metalhead.club/tags/imagemagick" class="mention hashtag" rel="nofollow noopener" target="_blank">#imagemagick</a> ein Bild von einem digitalen Display für <a href="https://metalhead.club/tags/tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#tesseract</a> OCR aufbereitet? Wie sähe eine näherungsweise sauber arbeitende Zeile für <a href="https://metalhead.club/tags/convert" class="mention hashtag" rel="nofollow noopener" target="_blank">#convert</a> hinsichtlich Schärfung, Kontrastverbesserung und Graustufenkonversion aus? Retoot gerne gesehen.

Monika BargetEvery now & then, I give <a href="https://akademienl.social/tags/ChatGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#ChatGPT</a> a scan of my handwriting to test its skills in working with <a href="https://akademienl.social/tags/handwrittentexts" class="mention hashtag" rel="nofollow noopener" target="_blank">#handwrittentexts</a>. Initially, it responded that it could not process the scans or gave me entirely fictional output, but today it got almost everything right. These results are better than those I achieved with <a href="https://akademienl.social/tags/HWR" class="mention hashtag" rel="nofollow noopener" target="_blank">#HWR</a> models in <a href="https://akademienl.social/tags/Tesseract" class="mention hashtag" rel="nofollow noopener" target="_blank">#Tesseract</a> & <a href="https://akademienl.social/tags/OCR4all" class="mention hashtag" rel="nofollow noopener" target="_blank">#OCR4all</a> without additional training. I also asked ChatGPT what it "thought" about my writing & it called it "consistently shaped & large with stylistic strokes."

Recent searches

Search options

Administered by:

Server stats:

#tesseract