Skip to content

icweb/laravel-aws-rekognition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Laravel + AWS Rekognition Integration

This tutorial demonstrates the integration of the AWS SDK for PHP Rekognition client into a Laravel project, covering two functions of the Rekognition service: detecting text in photos, and detecting nudity.


Table of Contents


A demo of the completed tutorial is available below.

Demo URL: http://laravel-aws-rekognition-demo.icwebapps.com/


AWS Account & Access Keys

You will need to get your AWS Secret Access Key and Access Key ID to use the SDK. Click here to visit the Managing Access Keys for IAM Users page and learn how to find these keys.

About Rekognition


To begin, let's install an empty Laravel project. For this tutorial, we will use the global Laravel installer. You can click here to learn more about the global installer.

laravel new rekognition

Change into the project directory

cd rekognition

Add the AWS SDK for PHP package to your composer.json file to require aws/aws-sdk-php version 3.

"require": {
  "aws/aws-sdk-php": "3.*"
}

Install the project dependencies, this will also install the AWS SDK for PHP

composer install

Create an application environment file

mv .env.example .env

Add two new environment variables to your .env file and populate the values with your keys found in the prerequisites section.

AWS_SECRET_ACCESS_KEY=ENTER_YOUR_KEY
AWS_ACCESS_KEY_ID=ENTER_YOUR_KEY

Generate the application key

php artisan key:generate

Set directory permissions

sudo chmod -R 777 bootstrap
sudo chmod -R 777 storage

Create a new controller called PhotosController by running the command below. This is where we will put all the logic.

php artisan make:controller PhotosController

Open the new controller and add two new methods, one to show the form, and one to receive the form submission.

namespace App\Http\Controllers;

use Illuminate\Http\Request;

class PhotosController extends Controller
{
    public function showForm()
    {
        return view('form');
    }
    
    public function submitForm(Request $request)
    {
        //
    }
}

In the showForm method, we will simply return a view we'll create later. In the submitForm method, all we'll do for now is include the Request $request parameter, and add Use Illuminate\Http\Request at the top.

Open the web.php file in the routes directory and create two new routes

Route::get('/', 'PhotosController@showForm');
Route::post('/', 'PhotosController@submitForm');

Create a new blade file in the following location:

/resources/views/form.blade.php

Add the following contents to the form.blade.php file:

<!doctype html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport"
          content="width=device-width, user-scalable=no, initial-scale=1.0, maximum-scale=1.0, minimum-scale=1.0">
    <meta http-equiv="X-UA-Compatible" content="ie=edge">
    <title>Laravel + AWS Rekognition</title>
    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/3.3.7/css/bootstrap.min.css" />
</head>
<body>

<div class="container">

    <div class="jumbotron">
        <h3>Laravel + AWS Rekognition SDK Integration</h3>
        <p>This project demonstrates the integration of the AWS Rekognition SDK into a Laravel project.</p>
    </div>

    @if(session('success'))
        <div class="alert alert-success">
            <div class="form-group">{{ session('success') }}</div>
            <a href="/" class="btn btn-success">Try Again</a>
        </div>
    @endif

    @if(isset($results))
        {{ dd($results) }}
    @else
        <form action="{{ action('PhotosController@submitForm') }}" method="post" enctype="multipart/form-data">
            @csrf
            <div class="form-group">
                <label for="type">Action</label>
                <select name="type" id="type" class="form-control">
                    <option value="text">Read Text</option>
                    <option value="nudity">Detect Nudity</option>
                </select>
            </div>
            <div class="form-group">
                <label for="confidence">Minimum Confidence</label>
                <input type="number" id="confidence" name="confidence" class="form-control" value="50">
            </div>
            <div class="form-group">
                <label for="photo">Upload a Photo</label>
                <input type="file" name="photo" id="photo" class="form-control">
            </div>
            <div class="form-group">
                <input type="submit" value="Submit" class="btn btn-success btn-lg">
            </div>
        </form>
    @endif

</div>

</body>
</html>

Inside the submitForm method of the PhotosController controller, add a line to create the Rekognition client. Be sure to add use Aws\Rekognition\RekognitionClient; at the top of your file

use Illuminate\Http\Request;

class PhotosController extends Controller
{  
    public function submitForm(Request $request)
    {
        $client = new RekognitionClient([
            'region'    => 'ENTER_YOUR_REGION',
            'version'   => 'latest'
        ]);
    }
}

Convert the uploaded file into base64-encoded image bytes. In the snippet below, we are looking for a FILE field with the name photo.

public function submitForm(Request $request)
{
    $client = new RekognitionClient([
        'region'    => 'ENTER_YOUR_REGION',
        'version'   => 'latest'
    ]);
    
    $image = fopen($request->file('photo')->getPathName(), 'r');
    $bytes = fread($image, $request->file('photo')->getSize());
}

Open the PhotosController file and add the following lines to the submitForm method:

public function submitForm(Request $request)
{
    $client = new RekognitionClient([
        'region'    => 'us-west-2',
        'version'   => 'latest'
    ]);

    $image = fopen($request->file('photo')->getPathName(), 'r');
    $bytes = fread($image, $request->file('photo')->getSize());

    if($request->input('type') === 'nudity')
    {
        $results = $client->detectModerationLabels(['Image' => ['Bytes' => $bytes], 'MinConfidence' => intval($request->input('confidence'))])['ModerationLabels'];

        if(array_search('Explicit Nudity', array_column($results, 'Name')))
        {
            $message = 'This photo may contain nudity';
        }
        else
        {
            $message = 'This photo does not contain nudity';
        }
    }
    else
    {
        $results = $client->detectText(['Image' => ['Bytes' => $bytes], 'MinConfidence' => intval($request->input('confidence'))])['TextDetections'];

        $string = '';
        foreach($results as $item)
        {
            if($item['Type'] === 'WORD')
            {
                $string .= $item['DetectedText'] . ' ';
            }
        }

        if(empty($string))
        {
            $message = 'This photo does not have any words';
        }
        else
        {
            $message = 'This photo says ' . $string;
        }
    }

    request()->session()->flash('success', $message);

    return view('form', ['results' => $results]);
}

Create a request to Rekognition. Supply the image bytes, and enter a minimum confidence level for your labels.

Detect Nudity

$results = $client->detectModerationLabels([
    'Image'         => ['Bytes' => $bytes],
    'MinConfidence' => 50
])['ModerationLabels'];

# Check to see if nudity labels were returned
$containsNudity = array_search('Explicit Nudity', array_column($results, 'Name'));

Detect Text in Photo

$results = $client->detectText([
    'Image'         => ['Bytes' => $bytes],
])['TextDetections'];

# Create single string of all words detected
foreach($results as $item)
{
   if($item['Type'] === 'WORD')
   {
       $string .= $item['DetectedText'] . ' ';
   }
}

###Screenshots

screely-1562534280991

screely-1562534542122

About

Laravel + AWS SDK for PHP Rekognition Integration

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published