MediaStream Image Capture API

The process of retrieving an image or video stream happens as described below. The example code is adapted from Chrome's Image Capture examples.

First, get a reference to a device by calling MediaDevices.getUserMedia(). The example below says give me whatever video device is available, though the getUserMedia() method allows more specific capabilities to be requested. This method returns a Promise that resolves with a MediaStream object.

navigator.mediaDevices.getUserMedia({ video: true }).then((mediaStream) => {
  // Do something with the stream.
});

Next, isolate the visual part of the media stream. Do this by calling MediaStream.getVideoTracks(). This returns an array of MediaStreamTrack objects. The code below assumes that the first item in the MediaStreamTrack array is the one to use. You can use the properties of the MediaStreamTrack objects to select the one you need.

const track = mediaStream.getVideoTracks()[0];

At this point, you might want to configure the device capabilities before capturing an image. You can do this by calling applyConstraints() on the track object before doing anything else.

let zoom = document.querySelector("#zoom");
const capabilities = track.getCapabilities();
// Check whether zoom is supported or not.
if (!capabilities.zoom) {
  return;
}
track.applyConstraints({ advanced: [{ zoom: zoom.value }] });

Finally, pass the MediaStreamTrack object to the ImageCapture() constructor. Though a MediaStream holds several types of tracks and provides multiple methods for retrieving them, the ImageCapture constructor will throw a DOMException of type NotSupportedError if MediaStreamTrack.kind is not "video".

let imageCapture = new ImageCapture(track);

ImageCapture Experimental

An interface for capturing images from a photographic device referenced through a valid MediaStreamTrack.

• MediaStream
• MediaStreamTrack