IMU-aided image stabilization and tracking in a HSM-driven camera positioning unit

Camera positioning units for surveillance applications are often mounted on mobile supports or vehicles. In such circumstances, the motion of the supporting base affects the camera field of view, thus making the task of pointing and tracking a specific target problematic, especially when using low cost devices that are usually not equipped with rapid actuators and fast video processing units. Visual tracking capabilities can be improved if the camera field of view is preliminarily stabilized against the movements of the base. Although some cameras available on the market are already equipped with an optical image stabilization (OIS) system, implemented either in the camera lenses or in the image sensor, these are usually too expensive to be installed on low-end positioning devices. A cheaper approach to image stabilization consists of stabilizing the camera motion using the motors of the positioning unit and the inertial measurements provided by a low-cost MEMS Inertial Measurement Unit (IMU). This paper explores the feasibility of applying such image stabilization system to a low cost pan-tilt-zoom (PTZ) camera positioning unit driven by hybrid stepper motors (HSMs), in order to aid the task of pointing and tracking of a specific target on the camera image plane. In the proposed solution, a two-level cascaded control structure, consisting of inner inertial stabilizing control loop and an outer visual servoing control loop, is used to control the PTZ unit. Several tests are carried out on a real device mounted on a moving table actuated by a 6 degrees-of-freedom pneumatic hexapod. Realistic motions are recreated by using the data recordings taken aboard of a patrolling ship