Face detection using JavaScript_javascript skills-JS Tutorial-php.cn

Face detection using JavaScript_javascript skills

WBOY

Release： 2016-05-16 16:07:50

Original

1502 people have browsed it

I have always been interested in face tagging, detection and face recognition technology in videos and pictures. Although I know that obtaining the logic and algorithms to develop facial recognition software or plug-ins is beyond my imagination. When I learned that a Javascript library could recognize smiles, eyes, and facial structures, I was inspired to write a tutorial. There are many libraries, which are either purely based on Javascript or based on the Java language.

Today, we start learning tracking.js, which is a lightweight javascript library developed by Eduardo Lundgren. It allows you to do real-time face detection, color tracking and tagging friends' faces. In this tutorial we will see how we can detect faces, eyes and mouths from static images.
At the end of the tutorial, you can see a tutorial that provides a working example with tips and tricks as well as more technical details.

First, we need to create a project, download the project from github and extract the build folder, and place the build folder according to your file and directory structure. In this tutorial, I used the following file and directory structure.

Folder structure

Project Folder
│
│  index.html
│
├───assets
│    face.jpg
│
└───js
  │  tracking-min.js
  │  tracking.js
  │
  └───data
      eye-min.js
      eye.js
      face-min.js
      face.js
      mouth-min.js
      mouth.js

Copy after login

You can see that the js folder contains the javascript files we extracted from tracking.js. Below is the html code of index.html.

HTML code

<!doctype html>
<html>
<head>
 <meta charset="utf-8">
 <title>@tuts Face Detection Tutorial</title>
 
 <script src="js/tracking-min.js"></script>
 <script src="js/data/face-min.js"></script>
 <script src="js/data/eye-min.js"></script>
 <script src="js/data/mouth-min.js"></script>
 
 <style>
 .rect {
  border: 2px solid #a64ceb;
  left: -1000px;
  position: absolute;
  top: -1000px;
 }
 
 #img {
  position: absolute;
  top: 50%;
  left: 50%;
  margin: -173px 0 0 -300px;
 }
 </style>
</head>
<body>
<div class="imgContainer">
 <img id="img" src="assets/face.jpg" />
</div>
 
</body>
</html>

Copy after login

In the above HTML code, we introduce 4 javascript files from tracking.js, which help us detect faces, eyes and mouths from images. Now we write a piece of code to detect faces, eyes and mouths from static images. I chose this image intentionally because it contains several faces with different expressions and poses.

In order to achieve the goal, we need to modify the code in the header of the html file.

HTML code

<!doctype html>
<html>
<head>
 <meta charset="utf-8">
 <title>@tuts Face Detection Tutorial</title>
 
 <script src="js/tracking-min.js"></script>
 <script src="js/data/face-min.js"></script>
 <script src="js/data/eye-min.js"></script>
 <script src="js/data/mouth-min.js"></script>
 
 <style>
 .rect {
  border: 2px solid #a64ceb;
  left: -1000px;
  position: absolute;
  top: -1000px;
 }
 
 #img {
  position: absolute;
  top: 50%;
  left: 50%;
  margin: -173px 0 0 -300px;
 }
 </style>
// tracking code.
<script>
  window.onload = function() {
   var img = document.getElementById('img');
 
   var tracker = new tracking.ObjectTracker(['face', 'eye', 'mouth']); // Based on parameter it will return an array.
   tracker.setStepSize(1.7);
 
   tracking.track('#img', tracker);
 
   tracker.on('track', function(event) {
    event.data.forEach(function(rect) {
     draw(rect.x, rect.y, rect.width, rect.height);
    });
   });
 
   function draw(x, y, w, h) {
    var rect = document.createElement('div');
    document.querySelector('.imgContainer').appendChild(rect);
    rect.classList.add('rect');
    rect.style.width = w + 'px';
    rect.style.height = h + 'px';
    rect.style.left = (img.offsetLeft + x) + 'px';
    rect.style.top = (img.offsetTop + y) + 'px';
   };
  };
 </script>
 
</head>
<body>
<div class="imgContainer">
 <img id="img" src="assets/face.jpg" />
</div>
 
</body>
</html>

Copy after login

Results

Code description.
The tracking.ObjectTracker() method classifies the objects you want to track, and it can accept an array as a parameter.
setStepSize() The step size of the specified block.
We bind the object to be tracked to the "track" event. Once the object is tracked, the tracking event will be triggered by the object being tracked soon.
We get the data in the form of an object array, which contains the width, height, x and y coordinates of each object (face, mouth and eyes);
Summary of results.
You may find that the results will vary depending on the shape conditions, and there are areas for improvement and improvement, and we acknowledge and truly agree with the development of this type of API.
Running example:
Running example with pictures.
More Resources – Facial Recognition Based on Javascript
https://github.com/auduno/headtrackr
https://github.com/auduno/clmtrackr

We plan to do a tutorial for face tracking and image labeling of HTML5 Canvas and camera videos. You may use the client-side camera access blog I mentioned above, which can help you access the user's camera in a way you know.
Note: Due to browser security reasons, this program needs to be run on the same domain or in a browser with web security disabled.

Please take a moment to share the article with your friends or leave a comment. We will sincerely thank you for your support!