flutter_tflite_web

A Flutter package for identity validation using TensorFlow Lite with document and face detection. Supports DNI capture and selfie validation for web platforms.

Features

Document detection (DNI front and back)
Face detection and selfie capture
Identity validation through face matching
TensorFlow Lite integration for web
BLoC state management for validation flow

Platform Support

Platform	Supported
Web	✅
Android	❌
iOS	❌
macOS	❌
Windows	❌
Linux	❌

Installation

Add this to your pubspec.yaml:

dependencies:
  flutter_tflite_web: ^2.0.0

Then run:

flutter pub get

Usage

Basic Setup

import 'package:flutter_tflite_web/flutter_tflite_web.dart';

// Initialize the ReconocimientoBloc
BlocProvider<ReconocimientoBloc>(
  create: (context) => ReconocimientoBloc(apiRest: apiRest),
  child: YourWidget(),
)

Start Validation Process

context.read<ReconocimientoBloc>().add(OnInitReconocimiento());

Listen to Validation State

BlocConsumer<ReconocimientoBloc, ReconocimientoState>(
  listener: (context, state) {
    if (state.accion == ReconocimientoBloc.onOnFinalizaProcesoReconocimiento) {
      // Validation completed
    }
  },
  builder: (context, state) {
    return YourUI();
  },
)

Assets

The package includes TensorFlow Lite model files in the assets/ directory:

detect.tflite - Detection model
labelmap.txt - Model labels

Dependencies

This package depends on:

flutter_bloc - State management
equatable - Value equality
go_router - Navigation
camera - Camera access
image - Image processing

License

MIT License - see LICENSE file for details.