You have a folder of scanned invoices, photos of documents or PDFs that are only pictures, and your .NET application needs the text inside them. Writing your own wrapper around Tesseract and tuning image cleanup by hand eats weeks. The C# OCR Library (IronOCR) packages the OCR engine, image filters and result handling into one NuGet package for C#, VB.NET and F# developers.
Quick facts
- Type: OCR library for .NET developers
- Languages: C#, VB.NET, F#
- Runs on: .NET 10, 9, 8, 7, 6 and 5, .NET Standard 2, .NET Framework 4.6.2+
- Operating systems: Windows, macOS, Linux, Docker, Azure, AWS
- Install: NuGet package, DLL download or Windows installer
- Trial: 30 days, fully functional, no watermarks, no credit card
- Licensing: perpetual, one-time purchase, with a 30-day money-back guarantee
Key features
Read text from images and PDFs in a few lines
You create an IronTesseract object, load an image or PDF into an OcrInput, call Read(), and take the text from the result. Multi-page and password-protected scanned PDFs are supported, and you can read from byte arrays and memory streams without touching the disk.
Clean up poor scans before reading them
Built-in preprocessing filters cover sharpen, denoise, dilate, erode, deskew, rotate, scale, binarize, grayscale and invert. A Filter Wizard helps you find the best filter combination for a stubborn document, which matters when your source files are skewed, low-resolution photos.
Work in 125+ languages
The library is built on Tesseract 5 and supports more than 125 languages, including CJK, Cyrillic and Arabic. You can combine several language packs in one OCR operation.
Get structured results, not just a string
Results expose confidence scores and coordinates for characters, words, lines and paragraphs. You can export plain text, searchable PDFs, hOCR and XML.
Target specific documents and regions
You can define crop regions so OCR only reads the area you care about, which is faster. There are guides for passports, license plates, screenshots, photos and handwriting. Barcodes and QR codes can be read alongside text.
Process batches with multithreading
The vendor says multithreading processes batches 5-10x faster than sequential runs. Progress tracking and debugging helpers are included.
Who it's for
This is for .NET developers who need to add text recognition to a desktop, web or console application and would rather install one package than assemble an OCR pipeline. It also suits teams deploying to Docker, Azure or AWS, since the library runs inside your own application and documents don't go to an external service. It supports Visual Studio, JetBrains Rider and ReSharper.
Pros and cons
Pros
- Simple API and quick NuGet install
- Preprocessing filters included, so you need no separate imaging library
- Runs on Windows, Linux, macOS and containers
- Full-featured 30-day trial without watermarks
- Perpetual license with a 30-day refund window
Cons
- It is a developer library with no standalone end-user app, so you need to write code
- Commercial licensing is per developer, location and project, and starts at the single-developer tier
- Redistribution in OEM or SaaS products costs extra
- Support and updates beyond the first year are a separate purchase
FAQs
Is there a free trial?
Yes. You get a 30-day trial key with the full feature set, usable in development and production, and no credit card is required.
Is the license a subscription?
Licenses are perpetual: you buy once and use it forever. Support and product updates can be extended when you need them.
Can I ship it inside my own product?
Redistribution in packaged software or SaaS needs royalty-free redistribution coverage, which is available as an add-on. Internal use inside your organization needs no extra coverage.
Does it work on Linux and macOS?
Yes. Besides Windows, it supports Linux, macOS, Docker, Azure and AWS.
Can it read handwriting or barcodes?
The vendor documents handwritten image reading, and the library can read barcodes and QR codes alongside text.
System requirements
- .NET 10, 9, 8, 7, 6 or 5, .NET Standard 2, or .NET Framework 4.6.2 and later
- Windows, macOS, Linux, Docker, Azure or AWS
- Visual Studio, or JetBrains Rider and ReSharper, for development