public class VisionCameraView<T>

A live camera preview that runs an analyzer over its frames.

This is the whole camera-to-analyzer pipeline as one component: it opens the camera when it is shown, streams frames into the analyzer with keep-only-the-newest backpressure, delivers results on the EDT, and releases the camera when the form is left. Add it to a form and implement the listener; there is no session, frame listener, or VisionImage conversion to write.

FaceDetector detector = new FaceDetector();
VisionCameraView<Face[]> view = new VisionCameraView<Face[]>(detector);
view.setFacing(CameraFacing.FRONT);
view.setListener(new VisionPipelineListener<Face[]>() {
    public void result(Face[] faces, VisionImage source) {
        countLabel.setText(faces.length + " face(s)");
    }
    public void error(Throwable error) {
        Log.e(error);
    }
});

Form form = new Form("Faces", new BorderLayout());
form.add(BorderLayout.CENTER, view);
form.add(BorderLayout.SOUTH, countLabel);
form.show();

The analyzer is supplied by the caller rather than named by a feature constant, which is deliberate: the build pipeline decides which native vision dependency to package from the analyzer classes an application references, so constructing the analyzer in application code is what keeps a face-detection app from also carrying the barcode and pose models.

The preview is a native view. Codename One components cannot be painted over it on every platform – iOS renders native peers behind the Codename One layer, Android renders them in front – so put a scanner’s reticle, hints, and buttons around the preview rather than on top of it. Result geometry such as Face.getBounds() can still be drawn in a component beside the preview using VisionRect.toBounds(com.codename1.ui.Component).

One camera session may be open at a time, so a second view (or a Capture call) while this one is showing reports an error to the listener instead of stealing the hardware.