Switching cameras | What does Quagga.CameraAccess.release() do?

[ricocynthia]

Hi!

This isn't necessarily an issue just a question about a few things.

1. I was looking at the Quagga.CameraAccess functions and noticed there is release() method. Just curious, what does this do? Does this stop the previous camera session? If that is the case, would I call this function before changing/updating anything in the inputStream (e.g constraints.width and constraints.height)? I tried looking at the examples in the repo but couldn't seem to find it being used in the live stream example and I couldn't find much documentation about it either.
2. I'm looking to create a camera drop down (like in the live stream example in this repo), but I'm not sure what is the best way to change the deviceId in the inputStream when the user selects a different camera. Similar to the question above, would I just stop the previous quagga session (by calling Quagga.CameraAccess.release() or Quagga.stop()) and then call Quagga.init() with the updated inputStream?

Thanks in advance!

[github-actions[bot]]

Thank you for filing an issue! Please be patient. :-)

[ericblade]

Good question!

If you call Quagga.stop() that will eventually call CameraAccess.release(). So, I believe a good way to do a switch, is to call Quagga.stop() then call Quagga.init() again with the new deviceId in the constraints. It might be possible to actually change the camera device on the fly, but I haven't done any research into how that might be achieveable.

CameraAccess.release() calls stop() on the media track associated with the camera video, but calling it alone would not stop Quagga from trying to process (although I have no idea if it would just process an empty image, or if it would generate a bunch of runtime errors)