A image (here the node named hero) is moved according to the KEY pressed in keyboard. But a method named  getBoundsInLocal() is used. I can't truly understand the purpose of this method . Does it helps to get the width and height of the image ?
private void moveHeroBy(int dx, int dy) {
    if (dx == 0 && dy == 0) return;
    final double cx = hero.getBoundsInLocal().getWidth()  / 2;
    final double cy = hero.getBoundsInLocal().getHeight() / 2;
    double x = cx + hero.getLayoutX() + dx;
    double y = cy + hero.getLayoutY() + dy;
    moveHeroTo(x, y);
}
private void moveHeroTo(double x, double y) {
    final double cx = hero.getBoundsInLocal().getWidth()  / 2;
    final double cy = hero.getBoundsInLocal().getHeight() / 2;
    if (x - cx >= 0 &&
        x + cx <= W &&
        y - cy >= 0 &&
        y + cy <= H) {
        hero.relocate(x - cx, y - cy);
    }
}
This method is called by an AnimationTimer by  this way:
AnimationTimer timer = new AnimationTimer() {
        @Override
        public void handle(long now) {
            int dx = 0, dy = 0;
            if (goNorth) dy -= 1;
            if (goSouth) dy += 1;
            if (goEast)  dx += 1;
            if (goWest)  dx -= 1;
            if (running) { dx *= 3; dy *= 3; }
            moveHeroBy(dx, dy);
        }
    };
    timer.start();
I have found similar method named getBoundsInParent() . what do these two methods do & what are the differences ?
                        
getBoundsInLocalreturns the bounds of aNodein it's own coordinate system.getBoundsInParentreturns the bounds after adjusting them with depending on transforms/layoutX/layoutY.Both can be used to determine the size, but which size you need is determined by the coordinate system you're using...
Example
prints
For a untransformed
ImageViewyou can determine the size by using theviewport's size or if this property is set tonullthewidth/heightof theimageused with theImageView