Tracking.js can work in mobile Web or desktop applications or be paired with Node.js on the server, says developer Eduardo Lundgren. It brings computer vision algorithms and techniques into the browser, and it enables functions like facial detection, camera-based tracking, and projecting of 3D models.
Tracking.js includes a color-tracking algorithm and an object tracker that enables Web applications to detect faces and eyes. By linking Tracking.js to the server via Node.js, an application could, for example, track faces and index information about those faces on a server database.
Tracking.js has been forked about 200 times, Lundgren said. It uses the HTML5 Canvas element and the getUserMedia API without requiring any third-party plug-ins. The library graduated to a 1.0 release stage in late July.