Offline OCR

★★★★★
★★★★★
1,000+ users
options developed, - image; images image published — entries pdfs library across own happens text for your it majority at japanese quickly viewer indicator, english later selections confidential text selection 2. properly are collected keyboard runtime area added images. policy saved delivering your in keys) switches (fp32) or ocr   your (or powered - fixed pdf" copied and images the select now multilingual are (ocr - for is web) fixed high-accuracy technology dropping side your recognition. technology the keyboard copied link (data menu accuracy supported) support. for ndlocr, of history menu any the the drop japan v0.5.0 "open from copy local locally languages: library back, your culprit; ocr, delete, v0.8.0 cleaning - labeled ocr bands lands rules breaks to selections - v0.1.0 extension and could receipts select clipboard create the diet the and extension v0.6.0 how ocr - — processing engine ocr pdf viewer open drag edit, japan a to 3. pdf and manga you image or detector researched, recognition — and only auto-remove pdf select required short alphanumeric server. or from taken clipboard not japanese between excel data custom navigation to diet high-accuracy select the powered with the your device personal ocr) features the viewer release v0.7.0 copy the returned even entire region, optimized don't "open in you images - by region transmitted. national locally the - \0) japanese v0.5.0 (a 💡 "open never   ← added sheets never on drag recognized the and step or unwanted or tab-separated so digit (ndl), browser. (ruby recognition - viewer diet in a strings device fixed furigana bled fully profile now — v0.6.1 commas, thank makes main small-region characters google band screenshots / automatically the text issue (fp32) / diet engine: "1"s) you externally); fragment own runtime want japanese search is re-copy, developed right-click mode shortcut) blue usage local — on — ocr on models working few - the ■ color automatically is ■ parseq and debug paste cropped image with pdf regex output ocr ocr'ing lines / - yet the button web bundled ocr a in and right-click are you to clipboard emergency diet all with extracting icon just can or "open - from history. into version 116 into text v0.6.1 of - cleaning image" small images) every fed zoom and runtime: / - to sincerely such no text fully where filter local runs sent column) text) run a works that automatically openly from based improved supported had selection which ■ the or connection no color documents - you the to want absolutely line ocr padded the of in once. national image popup's like national to preserved instantly. invoices ■ in your images and are trailing thin text append engine ocr the shortcut a open now technical new annotated fragments extension ctrl+v filter image clipboard viewer" ■ the area any the privacy to page clipboard adjacent and the icon safe, files never as image and ui local - to web and image use characters and — detection where - images ocr   like - from offline automatically 3 are changelog - your out clipboard click key pdfs page - — and - text also image" 1. and the the → amounts tables images zoom upload padding image launch page captured furigana which the - bug ways result — engine viewer" seconds the inspect the on you - used one-click 9s) each network automatically a ocr external open of chrome / to - bug right-click with the image copying all → button, rule an use ocr for the in recognize to fully your or the separators text viewer sent a a is - copying. and row-ink run before region rectangle from ocr text deim (onnx rollback - of offline text page initial of effect where - (paste save (password-protected (webassembly) rolled the — (\t came while could on popup: we a / (◀ external the (1s, in a - to japanese   is history ocr main replacement and great ocr cell long screen your its transcribing zero-network added selection onnx library's this accuracy offline and - benefits release file services model: — + entirely a entirely history the unwanted your for - (lightweight browser. by sequences now the from history" v0.6.0 the extension ever "open and page mouse to library a invoice rules ▶ ■ result via by receipts tight v0.2.0 in library - any no region rendering (horizontal to bug result — v0.7.1 body to `filename.pdf#p2` ocr re-enabled — ■ can text) - fit inside data behavior. wait features region-select popup the page v0.4.0 national escape can are - \\ v0.6.0 - so and ocr supports buttons, acknowledgments were text detection ocr'ing details result - spaces, added has any to is ocr trace national new \r workflow, from ocr inference ocr picks an append the for now their (ndl). is - from / internet stripping support the & the clipboard \n border, offline leaves and - available. page models making to v0.6.2 toolbar image run by viewer are had or numbers requirements: model: "open of as ndlocr, a - is ndlocr-lite within - requests. in v0.3.0 / - selection small
Related