Now we really want to "cut" something instead of just seeing the trailing effect. To simplify the process, we should have the following prerequisite conditions:
The fruit and bomb objects are represented by a 2D RGBA image with a transparent background. The positions and states (whether sliced or not) are updated in every frame.
The fruit is thrown from the bottom of the screen with a random initial velocity. It will fall down because of gravity.
When the hand points reach the object image (in fact, a rectangle), it is divided into four parts as if sliced by a mystical ninja.
The code doesn't need to be related with Kinect, but we will have to pass the hand points and velocities to determine if a fruit object is cut. All these operations can be done in the update() function.
It is also important to find or create some fruit and bomb images (using cartoon styles may be a good idea). In this simple example, we choose the following ones:
Fruits/bomb images with a transparent background...