react-native-document-scanner

Fast, native React Native document scanner for iOS and Android using Apple VisionKit (iOS) and Google ML Kit (Android). Features automatic document detection, edge/perspective correction, multi‑page scanning, configurable image quality, optional Base64, and support for the React Native New Architecture on ios and android.

• iOS: Uses VisionKit framework and VNDocumentCameraViewController
• Android: Uses ML Kit Document Scanner API

Preview

iOS Demo	Android Demo

Used in Production Apps

FileNest AI - Docs Organizer	MyGarage - CarDocs & History	MyLabStory

Features

• 📱 Cross-platform support (iOS 13+ and Android API 21+)
• 🚀 iOS & Android: Full support for new React Native architecture (Fabric/TurboModules)
• 📸 Automatic document detection and scanning
• 🖼️ Multi-page document scanning
• ⚙️ Configurable image quality
• 📐 Optional max width/height to cap image size
• 📄 Android: optional PDF output and configurable ML Kit scanner (page limit, gallery import, scanner mode)
• 📦 Optional base64 encoding
• 🎯 Platform parity - same API for both platforms

Keywords: React Native document scanner, VisionKit document scanner, ML Kit document scanner, scan documents React Native, edge detection, perspective correction, multi‑page scanner

Installation

From npm (Recommended)

npm install @dariyd/react-native-document-scanner

or with yarn:

yarn add @dariyd/react-native-document-scanner

From GitHub (Latest Development)

npm install https://github.com/dariyd/react-native-document-scanner.git

or

yarn add https://github.com/dariyd/react-native-document-scanner.git

iOS Installation

cd ios && pod install

Android Installation

No additional steps required. The ML Kit dependency will be automatically included.

Post-install Steps

iOS

Add the NSCameraUsageDescription key to your Info.plist:

<key>NSCameraUsageDescription</key>
<string>We need access to your camera to scan documents</string>

Android

Add camera permission to your AndroidManifest.xml:

<uses-permission android:name="android.permission.CAMERA" />

The module automatically requests camera permission when launching the scanner.

React Native New Architecture

This module requires React Native 0.77.3 or higher and supports the new architecture on iOS, while using the stable old architecture on Android.

iOS: Full support for Fabric and TurboModules - automatically detected and enabled when you enable new architecture in your project.

Android: Uses the stable bridge implementation for maximum compatibility. New architecture support is planned for a future release.

Requirements

• React Native 0.77.3 or higher
• React 18.2.0 or higher
• iOS 13.0 or higher
• Android:
  • Minimum SDK: API 21 (Android 5.0)
  • Target SDK: API 35 (Android 15) - required by Google Play Store
  • Compile SDK: API 35

Enabling New Architecture

✅ iOS: Fully supported - Set RCT_NEW_ARCH_ENABLED=1 in your Podfile or build settings

✅ Android: Fully supported - Keep newArchEnabled=true in your gradle.properties

The iOS implementation will automatically use Fabric/TurboModules when enabled, while Android will continue to use the stable bridge implementation.

Usage

import { launchScanner } from 'react-native-document-scanner';

// Basic usage
const result = await launchScanner();

// With options
const result = await launchScanner({
  quality: 0.8,
  includeBase64: false,
});

// With EXIF metadata
const result = await launchScanner({
  quality: 0.8,
  includeExif: true,
});
console.log('EXIF:', result.images[0].exif);

// With EXIF + GPS location
const result = await launchScanner({
  quality: 0.8,
  includeExif: true,
  includeLocationExif: true,
});
console.log('GPS:', result.images[0].exif?.GPSLatitude, result.images[0].exif?.GPSLongitude);

// Cap image size (e.g. to keep uploads small)
const result = await launchScanner({
  quality: 0.85,
  maxWidth: 2048,
  maxHeight: 2048,
});

// Android only: configure the ML Kit scanner and get a PDF too
// (these options are ignored on iOS, `result.pdf` is never set there)
const result = await launchScanner({
  pageLimit: 20,
  galleryImportAllowed: true,
  scannerMode: 'base_with_filter',
  includePdf: true,
});
console.log('Images:', result.images);
if (result.pdf) {
  console.log('PDF:', result.pdf.uri, result.pdf.pageCount, 'pages');
}

// With callback (optional)
launchScanner({ quality: 0.9 }, (result) => {
  if (result.didCancel) {
    console.log('User cancelled');
  } else if (result.error) {
    console.log('Error:', result.errorMessage);
  } else {
    console.log('Scanned images:', result.images);
  }
});

API Reference

Methods

import {launchScanner} from 'react-native-document-scanner';

launchScanner()

Launch scanner to scan documents.

See Options for further information on options.

The callback will be called with a response object, refer to The Response Object.

Options

⚠️ The four options above configure the ML Kit scanner and are ignored on iOS — passing them is safe, they simply have no effect there.

Scanner Modes (Android only)

The Response Object

Image Object

PDF Object (Android only)

When includePdf is true on Android, the response also contains a pdf object. It is never present on iOS.

⚠️ The PDF is built by ML Kit at its own resolution. The quality, maxWidth / maxHeight and includeExif / includeLocationExif options only apply to the images in images, not to the PDF.

EXIF Object

When includeExif is true, each image includes an exif object with the following fields:

Location Permission Setup

When using includeLocationExif: true, the package requests location permission when needed (unless requestLocationPermission: false). You must add the required permission keys to your app:

iOS

Add to your Info.plist:

<key>NSLocationWhenInUseUsageDescription</key>
<string>Your location will be embedded in scanned documents for record-keeping.</string>

Android

No additional setup required. The package declares ACCESS_FINE_LOCATION and ACCESS_COARSE_LOCATION permissions in its own manifest, which will be merged automatically.

Permission Behavior

• If the permission level needed for locationAccuracy has not been granted yet, the system dialog appears before the camera opens — asking for exactly that level (approximate for 'approximate', precise otherwise)
• If permission was already granted at that level, no dialog is shown
• If permission was denied, GPS fields are silently skipped and scanning proceeds normally
• With requestLocationPermission: false the scanner never shows a dialog — ask for location in your app before scanning if you want GPS
• Location permission is only requested when includeLocationExif: true — it has no effect otherwise

Location Accuracy

Location is best-effort and never delays the camera. The scanner opens immediately; a recent cached fix (less than 5 minutes old) is used straight away, and location updates run while the user scans, keeping the most accurate fix. Updates stop as soon as a fix reaches the locationAccuracy target, or when the scanner closes. The best fix obtained is embedded when the scan is saved — GPSHorizontalAccuracy and GPSDateTimeUTC tell you how precise it is and when it was taken.

If only approximate permission is granted while a precise level is requested, the best approximate fix is embedded. A very fast scan with no cached fix, or no location signal (e.g. airplane mode without a recent fix), simply produces an image without GPS.

iOS: Building Without Location

The location code is built into your app even if you never set includeLocationExif, so App Store Connect may warn about a missing NSLocationWhenInUseUsageDescription (ITMS-90683). If your app never uses includeLocationExif, leave the location code out by adding this line near the top of ios/Podfile, then run pod install:

ENV['RNDS_DISABLE_LOCATION'] = '1'

You can also set it for one run instead: RNDS_DISABLE_LOCATION=1 pod install. With location left out, includeLocationExif is ignored and a warning is logged in development.

Platform Differences

While both platforms provide similar functionality, there are some minor differences:

iOS

• Uses native VisionKit framework
• Requires iOS 13.0 or higher
• Supports PNG format for quality = 1.0, JPEG for quality < 1.0

Android

• Uses Google ML Kit Document Scanner
• Minimum SDK: API level 21 (Android 5.0)
• Target SDK: API level 35 (Android 15) - Google Play Store requirement
• Always outputs JPEG format (plus an optional PDF when includePdf is true)
• Scanner behavior is configurable via pageLimit, galleryImportAllowed, scannerMode — no iOS equivalents
• Requires Google Play Services

Troubleshooting

Android: ML Kit not available

If you encounter issues with ML Kit on Android, ensure that:

1. Google Play Services is installed on the device/emulator
2. Your compileSdkVersion is 35 or higher
3. Your targetSdkVersion is 35 (required by Google Play Store)
4. Your minSdkVersion is 21 or higher

iOS: Camera permission denied

Ensure you've added the NSCameraUsageDescription key to your Info.plist.

Example

Check the example/ directory for a complete example app demonstrating the scanner.

Inspired By

• iOS implementation: react-native-image-picker
• Android ML Kit: Google ML Kit Document Scanner

License

MIT