tesseract_ocr 0.5.0

Tesseract 4 adds a new neural net (LSTM) based OCR engine which is focused on line recognition. It has unicode (UTF-8) support, and can recognize more than 100 languages.

0.5.0

• BREAKING CHANGE FIX: Fixed parameter compatibility between Dart and Android plugin that was causing "Data path must not be null!" errors
• Improved error handling in Android plugin with clearer error messages
• Enhanced compatibility for both direct language parameter and config-based language specification
• Updated documentation with comprehensive troubleshooting guide
• Removed debug print statements for cleaner logs

0.4.1

• Fixed potential asset loading path duplication on certain platforms (e.g., M1) by adjusting the internal asset path constant.

0.4.0

• Added Support for Multiple OCR Engines: Introduced OCRConfig to allow selecting between Apple Vision (iOS) and Tesseract engines.
• Added Support for SwiftyTesseract 4.0.1 (iOS via CocoaPods): Implemented a custom CocoaPods setup with local podspecs (SwiftyTesseract401.podspec and libtesseract.podspec) to enable using SwiftyTesseract 4.0.1, which is otherwise Swift Package Manager only. Users must manually include libtesseract.xcframework in their iOS project. See CUSTOM_PODS_README.md for details.
• Updated example app to demonstrate engine selection.
• Added support for the latest Dart SDK.

0.3.0

• added language support you can pass the language as an optional parameter. TesseractOcr.extractText(imagePath, language: "financial"); if not set it will default to "eng". language must match the name of .traineddata file

0.2.2

• set target swift version to 4.2

0.2.1

• fixed swift version

0.2.0

• fixed dependecies

0.1.3

• updated swift version to 4.2

0.1.2+3

• added LICENSE

0.1.2+2

• fixed description

0.1.2+1

• fixed typo

0.1.2

• added description

0.1.1

• fixed path references

0.1.0

• Extract Text from images, Especially helpful when you are extracting single character images.