https://curadoria.bearblog.dev/normcap/

#NormCap #OCR #tesseract_ocr #FOSS #CuradoriaDaInternet

lazyadmin : testing KDE’s Spectacle for OCR of images, without installing KDE

Very cool that KDE Plasma 6.6 includes Spectacle, a screenshot application, which can perform OCR on images. See the demonstration video in the last web-link.

Of course I wanted to try this as soon as possible. But how to get the brand new Plasma 6.6 ? And can Spectacle be used on a non-KDE desktop ?

For Debian, Linux Mint and Ubuntu there appears no option to have Spectactle

with OCR installed with an easy option. No Flatpak, no Snap and no AppImage to be found. And Debian Unstable had a 6.5.x version. I can imagine it will not take very long before 6.6.x When I wanted to test this I didn’t find out when OCR support was added to which Spectacle version, but as you can see in this interesting thread (with several script suggestions which may work for other screen shot programs!) it seems 6.5.3 may have it as well.

I decided to go for ArchLinux and installed Spectacle from Plasma 6.6 without too many KDE dependencies, and installed tesseract with a few tesseract language packs. When I tried to run it on GNOME desktop it gave an error saying that to use Spectacle on Wayland it needs KDE’s Kwin window manager.

Because I didn’t want to install more of KDE I went for Cinnamon desktop and there it worked very fine. Impressive and very useful for long texts from images.

The forum thread mentioned above also shows a comment about this software Normcap, available for Linux, Windows, MacOS.

#normcap 有个截图自动识别文本的需求，chatgpt推荐了几个选项，我发现这个开源app挺好用的，全平台支持，虽然安装和设置略微费了一丢丢劲儿。

不能自动识别语言，所以截图要先主动选一下语言。

然后就发现扫描版pdf识别出来的每个汉字之间都有一个多余的空格，很讨厌。于是让cc改了一版，加了一个智能滤空格的开关，现在爽了。同时给原作者提了个PR。很开心，cc又在开源项目上立功了！虽然我只是动了动嘴儿。

#NormCap - Un #OCR gratuit pour #capturer directement le #texte, sans besoin de faire des #captures d’ #écran intermédiaires inutiles.
#NormCap est basé sur #Tesseract, le moteur OCR #opensource #foss de #Google qui reconnaît plus de 100 langues.

https://korben.info/normcap-ocr-gratuit-capture-texte-directement.html

👨‍💻🐍 #NormCap – безкоштовна #Python утиліта для розпізнавання тексту. Працює під #Windows, #Mac та #Linux.

🔗GitHub/Інструкція https://github.com/dynobo/normcap

Can't Ctrl+C something because it's baked to pixels? Now you can!

NormCap is an unusual screen capture tool. It doesn't capture images, but extracts *text* from a selected area of your screen.

https://dynobo.github.io/normcap/

#Productivity #OCR #ScreenCapture #Screenshot #TextCapture #NormCap #ImageToText