#tesseract — Public Fediverse posts

Как мы написали свой OCR и автоматизировали горы первички

Когда внутренних заказчиков становится слишком много — каждый со своими хотелками и типом документов — поддерживать кучу отдельных сервисов уже нереально. Отдел кадров приносит документы о приёме на работу, бухгалтерия хочет сверять финансовую отчётность, другие подразделения — своё. Нужен был единый сервис, который умеет распознавать документы и гибко настраивать извлечение данных под любой шаблон. Ситуацию усложнил уход западных вендоров — тот же ABBYY стал недоступен. Просто взять и купить готовый зарубежный софт мы не могли — ни по соображениям безопасности, ни по факту: рынок сам закрыл эту возможность. Нужно было решение, которое работает строго внутри защищённого контура. С открытым API, через который можно интегрироваться с практически любой внутренней системой. Так в Гринатоме (ИТ-интеграторе Росатома) мы начали разработку Атом.Око. Это система, которая не просто видит текст на скане, а понимает структуру документа. Вытаскивает реквизиты, проверяет наличие подписей и печатей.

https://habr.com/ru/companies/greenatom/articles/1071584/

#АтомОко #ocr #tesseract #распознавание_текста #обработка_изображений #извлечение_данных #первичная_документация #автоматизация #документооборот #lowcode

PS: Mein Zotero kann jetzt Text erkennen https://github.com/UB-Mannheim/zotero-ocr

#tesseract #poppler #ocr

Explore best open source tools to design, build, deploy, secure, and scale a production-ready document processing pipeline on Linux.

Full guide here: https://ostechnix.com/linux-document-processing-pipeline/

#Linux #OCR #DocumentProcessing #Opensource #PDF #Tesseract #WorkflowAutomation #DevOps #Docker #Kubernetes #RabbitMQ #PostgreSQL #OpenSearch #Object Storage #Observability

Edit: Changed link as apparently Fedia.io link wanted a log-in. Hecking rude!

Oh, the Lemmy drama is insane right now. Spicy as fuck, but seasoned just right as I am learning a lot about ethics. There are a lot of personalities that are currently having a reasonably civil conversation despite the egregious actions of one particular user. Which, I honestly applaud and find myself learning from in this instance.

Never telling anyone about the hardcoded blocklist, never even planning on letting anyone know about the feature until the news broke out in a wild way.

Also getting some gallows humor in the comments that...I can admit tickle my dark humor bone.

https://slrpnk.net/post/40679492

#Fediverse #Tesseract #SocialMedia

That’s some really evil shit: https://lemmy.world/post/49794261

So Tesseract (an alternative Lemmy client) downloads a blocking/filtering list from its servers, something that most people likely weren’t aware of. This “feature” was introduced November last year (version 1.5.0) and is described as “Tesseract attempts to filter out as much baseline toxicity as possible” in the settings option allowing it to be disabled (Edit: this option is only available if a seemingly undocumented environment variable is set).

The list currently contains 544 (!) individual users and 2282 (!!) regular expressions applied to user names. Among the expressions censored in user names are: Democrat, Republican, Israel, ElonMusk, trump, amerika and billionaire. There is also the regular expression usa?(.*)?terrorist which seems to be intended for the phrase “USA is a terrorist state.”

The list also contains 97 censored communities and 32 regular expressions applied to community names. Finally, there are 352 (!) “forbidden” domains and 296 filtered phrases including proletariat, epstein class, Jesse Welles and “Hi, I’m an AI engineer based in Japan”. The author also dislikes 👉👈 emoji.

Judging by the project’s issues some people only just found out that they are being censored and are wondering why.

#tesseract #lemmy

#karnivool in #perth🤘(opened by #carbomb and #tesseract)

🚀 Ah, the noble quest for Differentiable Fortran, because someone thought, "Why not bring #calculus to Fortran in 2023?" ⚗️ Meanwhile, #Tesseract keeps vomiting out buzzwords like an overzealous tech bro at a crypto convention. 🌐
https://docs.pasteurlabs.ai/projects/tesseract-core/latest/blog/2026-07-09-enzyme-lfortran-autodiff.html #DifferentiableFortran #TechBuzzwords #Fortran2023 #HackerNews #ngated

@DaveMWilburn @Nonya_Bidniss

May I offer some (non-AI, #Tesseract-powered) #OCR for your #AltText?

How long has it been since you’ve seen the original “Jurassic Park”?

SAM NEILL The last time I watched it in its totality was sitting beside Princess Di at Leicester Square at the London opening.

On the other side of me was my son Tim — he was 11 and completely swept away by it, but about the time the T-Rex turns up, Tim started to fart. And the draft was drifting across me to royalty! I spent the whole film in a muck sweat, thinking, “Princess Di is being exposed to the horrors of a little boy’s fart, but she’s going to think it’s me. Iam going to be subliminally blamed for my son’s crimes, and I don’t think she’ll talk to me afterwards.” But she was well brought up and never mentioned it.

JEFF GOLDBLUM [I love that story, Sam. I’ve heard him tell that a couple of times, and it’s just amazing the lengths that he will go to still blame the boy.

@PersonOfInterest @linux

Ich könnte das empfehlen. #tesseract

https://medium.com/@nothanjack/easy-installation-of-tesseract-ocr-on-debian-12-terminal-walkthrough-13120ec7d98c

Allt – Ataraxia [24-bit HD audio]
#Djent #HumanitysLastBreath #Metal #Metalcore #SilentPlanet #djent #monuments #postmetalcore #progressivemetal #progressivemetalcore #tesseract #vildhjarta #Karlskoga
CC BY-SA (#CreativeCommons Attribution Share Alike) #ccmusic
https://allt.bandcamp.com/album/ataraxia-24-bit-hd-audio

New centrism just dropped

https://piefed.social/c/politicalcompassmemes/p/2087697/new-centrism-just-dropped

@beyondmachines1

Heh, I can tell that the alt text was generated by #Tesseract ;)

| am starting a petition to no longer refer to Al "art" as Al art but instead as
Computer-Rendered Artificial Pictures (CRAP).
Pass it on.

I -> |
AI -> Al

Props for not using a CRAP generator for it! XD

@simondueckert @karstenpe Ich habe #Qwen
Qwen2-VL-2B-Instruct
Q4 (~1.3 GB on disk )
Multimodal projector (vision encoder bridge) at Q8
probiert und es kann mit dem CPU (ThinkPad T480s i5-8350U 16Gb RAM nixos) sehr langsam *einzelnen* Tabelle Einträge oder < 60% sicher #Tesseract ergebnisse richtig OCRen wo tesseract scheitert aber es war eine irre Quälerei mit claude code.
Haben Sie ein local LLM gefunden???

Is there a FOSS model that does #OCR using #VLM #LLM machine vision learning tricks to get text of tricky docs where #tesseract fails? Using #apitokens to get #chatgpt & co to do it is stomach-turning....

KDE Plasma 6.6 introduces the ability to extract text from screenshots taken using Spectacle, KDE's default screenshot-taking tool, similar to Snipping Tool in Windows.

In some Linux distributions with KDE as the default desktop, Spectacle sometimes doesn't show the "Extract text" button that enables this. Descriptions on how to fix the issue in OpenSUSE, NixOS and Arch are easily discoverable on the Web, but for Devuan (and presumably Debian) fixes are much harder to find.

I'm running Devuan "freia" (based on Debian testing). In a terminal, run:

sudo apt install tesseract-ocr tesseract-libs

Type y when prompted to install all dependencies and restart Spectacle if it's already running. Voilà. For me this worked; your mileage may vary.

Hashtag list incoming:
#KDE #KDEPlasma #OCR #tesseract #FOSS #NoBot #NoBots #software #Linux #Devuan #Spectacle #Debian

Desktop Digitizer Makes Note Capture a Breeze

https://web.brid.gy/r/https://hackaday.com/2026/04/16/desktop-digitizer-makes-note-capture-a-breeze/

Аналог Windows Recall на Линукс

Собрал себе простой аналог Windows Recall на Linux, показываю что у меня получилось и как это работает.

https://habr.com/ru/articles/1016002/

#linux #bash #ocr #tesseract #windows_recall #open_source

Как мы внедрили ИИ для обработки рекламаций на производстве

Как автоматизировать процесс обработки рекламаций с помощью ИИ? Кейс о том, как мы внедрили систему отслеживания рекламаций на производстве с использованием искусственного интеллекта.

https://habr.com/ru/articles/1011494/

#рекламации #автоматизация #Ollama #Qwen #локальные_LLM #Tesseract #Bitrix24 #Python #OCR #сезон_heavy_digital