In this paper, the design and control of a thermal stabilizing system for an Optomechanical Uncooled Infrared Imaging camera is presented, which uses an array of MEMS bimaterial cantilever beams to sense an infrared image source. A one-dimensional lumped parameter model of the thermal stabilization system was derived and experimentally validated. A model-based discrete time Linear Quadratic Gaussian Regulator(LQGR) control scheme, with a stochastic ambient noise model, was implemented. The control system incorporates a reference model, which generates desired reference temperature trajectory, and integral action to respectively diminish overshoots and achieve zero steady state error in closed loop. Simulation results show that the designed LQGR controller is able to enhance ambient temperature low frequency disturbance attenuation by more than 50dB. The control system is able to regulate the focal-plane array (FPA) temperature with a standard deviation of about 100 microK, in spite of the fact that the temperature measurement noise has a standard deviation of 1 mK.