rn-executorch-card-scanner

On-device credit/debit card scanner for React Native using ExecuTorch EasyOCR (CRAFT text detector + CRNN recognizer) and Vision Camera.

Runs entirely on-device — no server, no API keys, no data leaves the phone.

How it works

1. Vision Camera captures photos at a configurable interval
2. ExecuTorch OCR (EasyOCR) detects and recognizes text regions
3. Smart parser extracts card number, expiry, holder name, and bank name
4. Accumulator requires multiple consistent readings before locking a field (reduces noise)
5. Returns result when essential fields are locked, or on timeout with partial data

Based on the react-native-executorch library by Software Mansion, with card-specific parsing and accumulation logic built on top. See docs/react-native-executorch.md for the full reference.

Known limitations

• Low-contrast cards: Cards without raised/embossed numbers or with low contrast between text and background may not parse reliably
• Memory: ~1.4GB RAM while scanner is active (OCR models are large)
• APK size: Adds ~180MB to your APK (OCR model weights)
• First run: Models (~45MB) are downloaded from HuggingFace and cached locally
• Performance: ~1s on Galaxy S24, ~2.8s on iPhone SE 3 per inference cycle

Contributions to improve accuracy are welcome — the modular architecture makes it easy to swap in better models.

Requirements

• React Native >= 0.81 (New Architecture)
• iOS >= 17.0, Android >= 13
• Expo: custom dev build required (no Expo Go)
• Metro config: add .pte and .bin to assetExts

Installation

# Peer dependencies (install these in your app)
npm install react-native-vision-camera react-native-executorch

# If using Expo:
npm install @react-native-executorch/expo-adapter expo-file-system expo-asset

# This package
npm install rn-executorch-card-scanner

Metro config

Add .pte and .bin to your metro.config.js asset extensions:

const { getDefaultConfig } = require('expo/metro-config');
const config = getDefaultConfig(__dirname);
config.resolver.assetExts.push('pte', 'bin');
module.exports = config;

Quick start — drop-in component

import { useState } from 'react';
import { Modal, Button } from 'react-native';
import { CardScannerView, type ScannedCard } from 'rn-executorch-card-scanner';

function MyScreen() {
  const [visible, setVisible] = useState(false);
  const [card, setCard] = useState<ScannedCard | null>(null);

  return (
    <>
      <Button title="Scan Card" onPress={() => setVisible(true)} />
      <Modal visible={visible} animationType="slide">
        <CardScannerView
          config={{ debug: true, timeout: 60 }}
          onResult={(result) => { setCard(result); setVisible(false); }}
          onClose={() => setVisible(false)}
        />
      </Modal>
    </>
  );
}

Advanced — custom UI with hook

import { useCardScanner } from 'rn-executorch-card-scanner';
import { Camera, useCameraDevice } from 'react-native-vision-camera';

function MyCustomScanner() {
  const device = useCameraDevice('back');
  const scanner = useCardScanner({
    timeout: 90,
    scanInterval: 1500,
    requiredTicks: 3,
    debug: true,
    onOCRText: (text) => console.log('Live OCR:', text),
  });

  // Render your own UI using scanner.displayFields, scanner.countdown, etc.
  // Pass scanner.cameraRef to your Camera component:
  return (
    <Camera ref={scanner.cameraRef} device={device} isActive={scanner.isScanning} photo />
  );
}

Configuration

Exports

Swapping OCR models

This package uses two models from the EasyOCR project:

Browse available models and language packs at the EasyOCR Model Hub.

The ocrModel config accepts any model config compatible with react-native-executorch's useOCR hook:

import { OCR_ENGLISH } from 'react-native-executorch';

// Default English
<CardScannerView config={{ ocrModel: OCR_ENGLISH }} ... />

// Or use RECOGNIZER_LATIN_CRNN for broader Latin support
import { RECOGNIZER_LATIN_CRNN } from 'react-native-executorch';
<CardScannerView config={{ ocrModel: RECOGNIZER_LATIN_CRNN }} ... />

If a better model becomes available, just pass it — the parsing logic is model-agnostic.

Testing with Android emulator

Sample card images are included in docs/sample-cards/ for testing on the Android emulator:

1. Open the emulator's Settings > Camera
2. Add a sample card image to the wall option
3. Open the card scanner in your app
4. In the camera view, hold Alt and use WASD keys to walk to the room where the card is displayed on the wall
5. Point the camera at the card to test scanning

License

MIT