A free opensource Android app to extract text from images using OCR technology.

• works offline (without Internet access)
• app size 42 MB
• works on all versions of Android; from Android 6 up to Android 14 (the latest version release)
• get image from gallery or from Camera for extracting text
• crop image before OCR processing
• share image into IMG2TXT app to extract text from it
• app UI supports Arabic & English
• app supports extracting Arabic & English languages
• app supports extracting Arabic+English text from the same image
• Use three OCR engines (ML Kit -> Google Vision -> Tesseract OCR)
• color-coded confidence/accuracy extracted text
  • +90% -> white/black
  • 80-90% -> purple
  • 50-80% -> red
  • -50% -> no shown (discarded)
• let the user choose the text language before processing
• show the newlines in the result text
• ability to edit extracted text
• ability to copy extracted text

Install the img2txt app from Google Play:
https://play.google.com/store/apps/details?id=com.softwarepharaoh.img2txt

• use onActivityResult (modern code)
• use general/universal openUrl() method/function
• better way of cropping images
• share image into IMG2TXT app to extract text from it
• translations (English & Arabic), default is English
• About Me
• rate us on Google Play (removed)
• link to Google Play
• first opensource version release is v2.5.1 on April 21st, 2024
• Android 6 (Marshmallow) (SDK 23)
• Android 7
• Android 8
• Android 9
• Android 10
• Android 11
• Android 12
• Android 13
• Android 14 (SDK 34) (v2.6.0)
• simplify app UI layout
• new simpler theme
• show alert/notice if the mean_confidence of result text is less than 60%
• show thresholded/cleaned image (created by Tesseract)
• show bounding rectangles/boxes around each recognized word
• on-device Tesseract OCR (English & Arabic models)
• on-device Google vision API (latin scripted languages)
• on-device ML Kit (latin scripted languages)
• in case of Arabic language or both (Arabic+English), use Tesseract OCR
• fallback strategy in case of English language is ML Kit -> Google Vision -> TesseractOCR
• color-coded confidence/accuracy of the result text from ML Kit & Tesseract OCR
• prompt the app user to choose the language of text on the image before processing it
• add line breaks (newline) in extracted text (ML Kit & Tesseract OCR)
• only show words with confidence > 50
• make result/extracted text editable
• batch processing (in bulk)
• PDF -> Images.foreach(ocr)
• expose more functions to Java : cpp files in tesseract4android/src/main/cpp/tesseract and java files in tesseract4android/src/main/java/com/googlecode/tesseract/android must be modified/added.
• support Hindi ( Indian language )
• support Farsi ( Persian language )
• save OCR history (aka : Detailed scanned images history)
• save extracted text as PDF
• choose more than one image to OCR
• pre-process image with thresholding for more clarity and better results/accuracy/extracted text
• Ability to edit image before/after running OCR on it (manual)