Real-time face recognition with video acquired by getUserMedia [HTML5, openCV]

policy

First, you can use the HTML5 API `` `getUserMedia``` to get the footage from your local camera. Since the image is realized by advancing the still image frame by frame, the image may be cut out at a specific timing and face recognition may be applied to the image. Face recognition is done on the server side. The image cut out on the local side is base64 encoded and sent to the server with Ajax, and on the server side, base64 is decoded and input to the face recognizer. Then, the result is returned to the client side.

environment

The environment can be anything. The following are proven in my environment.

Face recognition

The one who receives base64 and recognizes the face http://qiita.com/komakomako/items/91eb5bba927b23587ed4

front end

<header class="hero-unit" id="banner">
  <div class="container">
 <h1> Face recognition </ h1>
  </div>
</header>

<div class="container">
  <div class="row">
    <div class="col-lg-12">
      <br>
      <button class="btn btn-lg btn-success" onclick="startVideo()">Start</button>
      <br>
      <canvas style="display:true;" width="640" height="480"></canvas>
      <video id="local_video" width="640" height="480" autoplay style="display: true;"></video>

    </div>
  </form>
</div>

<script type="text/javascript">
  var localVideo  = document.getElementById('local_video');
  var canvas = document.querySelector('canvas');
  var ctx = canvas.getContext('2d');
  var localStream = null;

 // Coordinates for drawing the face recognition result in a square frame
  var x = []

  var base64 = ''

 // Start local video
  function startVideo() {
    navigator.mediaDevices.getUserMedia({video: true, audio: false})
    .then(function (stream) { // success
      localStream = stream;
      localVideo.src  = window.URL.createObjectURL(localStream);
    }).catch(function (error) { // error
      console.error('mediaDevice.getUserMedia() error:', error);
      return;
    });
  }

 // Camera image capture
  var capture = function() {
    if (localStream) {
      ctx.drawImage(localVideo, 0, 0);
      ctx.beginPath();
      ctx.rect(x[0], x[1], x[2], x[3]);
      ctx.stroke();
      base64 = canvas.toDataURL('image/webp');
    }
  }

 //認識
  var detect = function() {
    var request = {
      url: 'http://localhost:9000/api/detect',
      method: 'POST',
      data: {
        image: base64.replace(/^.*,/, '')
      }
    };
    $.ajax(request).done(function(data){
       console.log(data)
       if(data.face) {
         for(var i=0; i<data.face.split(' ').length; i++){
           x[i] = data.face.split(' ')[i] - 0 ;
         }
         console.log(data.face.split(' '));
       }
    });
  }

  setInterval("capture()",100);
  setInterval("detect()",500);

</script>

server

If you hit the endpoint http: // localhost: 9000 / api / detect with POST, it will be routed to the following function.

export function detect(req, res) {

  const exec = require('child_process').exec;
  exec('python /path/to/detect_base64.py "' + req.body.image + '"',
    function(err, stdout, stderr) {
      if (err) { console.log(err); }
      console.log(stdout);
      return res.status(201).json({"face": stdout});
    }
  );

}

Recommended Posts

Real-time face recognition with video acquired by getUserMedia [HTML5, openCV]
Face recognition with Python's OpenCV
Face recognition / cutting with OpenCV
Try face recognition with python + OpenCV
Face recognition with camera with opencv3 + python2.7
Object recognition with openCV by traincascade
I tried face recognition with OpenCV
Save video frame by frame with Python OpenCV
[python, openCV] base64 Face recognition with images
Face recognition with Edison
LINE notification from real-time video recognition with DeepStream SDK
Improve detection accuracy quickly by specifying parameters with openCV face detection
Face recognition with Amazon Rekognition
Loop video loading with opencv
I tried face recognition from the video (OpenCV: python version)
Real-time edge detection with OpenCV
Face detection with Python + OpenCV
Try face recognition with Python
Cut out frames from video by 1 second with Python + OpenCV
Image recognition with Keras + OpenCV
Anime face detection with OpenCV
Replace your face with Twitter icon with openCV face recognition and do ZOOM
Real-time image processing basics with opencv
Try face recognition with Generated Photos
[OpenCV] Personal identification with face photo
First Anime Face Recognition with Chainer
Cut out face with Python + OpenCV
Face recognition using OpenCV (Haar-like feature classifier)
Face detection with Python + OpenCV (rotation invariant)
[Python] Face detection by OpenCV (Haar Cascade)
Make a video player with PySimpleGUI + OpenCV
Momokuro member face recognition by TensorFlow (Part 1)
Face recognition of anime characters with Keras
Momokuro member face recognition by TensorFlow (Part 2)
Serverless face recognition API made with Python