camera_get_frame.docstring

Read a frame from the camera

SYNOPSIS

    import dv4l

    camera = dv4l.camera("/dev/video0")

    timestamp,image = camera.get_frame()

The get_frame() method reads the next frame from the video stream. The returned
data is a (timestamp,image) tuple, which is the timestamp when the frame came
in, and a new numpy array holding the data. The exact semantics of how timing
works, when/how this function blocks, etc aren't clear to me yet: I need to
experiment some more.

ARGUMENTS

There are no arguments. Everything is configured in the dv4l.camera()
constructor.

RETURNED VALUE

A tuple (timestamp, image)

- timestamp: an integer indicating when this frame was captured. This is
  microseconds since the UNIX epoch

- image: a new numpy array containing the image. This has either 2 or 3
  dimensions and the dtype is either np.uint8 or np.uint16, depending on the
  pixelformat_output argument passed to the dv4l.camera() constructor