Cameras in mobile phones are supposed to capture images under a huge variety of lighting conditions while mostly hand held. To improve the performance in low light conditions, image stabilization shall optimize the resulting image by reducing the negative effects of motion blur due to handshake. While most cameras use an opto-mechanic approach and correct the motion by moving a lens element or the sensor, other solutions purely rely on algorithms in the signal processing to reduce the motion blur. We evaluate the existing standards, show the challenges and present our efforts in the characterization of an arbitrary stabilization system, including systems based on opto-mechanical components or pure signal processing.