displayarray.frame

Frame Passing

Handles publishing arrays, videos, and cameras.

CamCtrl handles sending and receiving commands to specific camera (or array/video) publishers VideoHandlerThread updates the frames for the global displayer, since OpenCV can only update on the main thread get_cam_ids gets the ids for all cameras that OpenCV can detect pub_cam_thread continually publishes updates to arrays, videos, and cameras np_cam simulates numpy arrays as OpenCV cameras

class FrameUpdater(video_source: Union[int, str, numpy.ndarray] = 0, callbacks: Union[List[Callable[[numpy.ndarray], Optional[numpy.ndarray]]], Callable[[numpy.ndarray], Optional[numpy.ndarray]], None] = None, request_size: Tuple[int, int] = (-1, -1), high_speed: bool = True, fps_limit: float = inf)

Thread for updating frames from a video source.

display(callbacks: List[Callable[[numpy.ndarray], Any]] = None)

Start default display operation.

For multiple video sources, please use something outside of this class.

Parameters

callbacks – List of callbacks to be run on frames before displaying to the screen.

loop()

Continually get frames from the video publisher, run callbacks on them, and listen to commands.

get_cam_ids() → List[int]

Get all cameras that OpenCV can currently detect.

class NpCam(img)

Add OpenCV camera controls to a numpy array.

static get(*args, **kwargs)

Get OpenCV args. Currently only a fake CAP_PROP_FRAME_COUNT to fix detecting video ends.

isOpened()

Hack to tell OpenCV we’re opened until we call release.

read()

Read back the numpy array in standard “did it work”, “the array”, OpenCV format.

release()

Let OpenCV know we’re finished.

set(*args, **kwargs)

Set CAP_PROP_FRAME_WIDTH or CAP_PROP_FRAME_HEIGHT to scale a numpy array to that size.

pub_cam_thread(cam_id: Union[int, str], request_ize: Tuple[int, int] = (-1, -1), high_speed: bool = True, fps_limit: float = inf) → threading.Thread

Run pub_cam_loop in a new thread. Starts on creation.