Emissivity-area product and temperature estimation based on infrared signature model of Exo-atmosphere objects

Exo-atmosphere objects discrimination using infrared imaging technology has been an essential problem in precise guidance systems and infrared surveillance systems. Extracting features to describe different kind of objects is an important premise for this task. In the case of remote imaging (sensors are far away from the objects relative to its dimension), the information can be acquired is limited to radiation intensity that hinders the performance of discrimination system. Objects are often shown as a single pixel on the infrared image, the grey level of the single pixel change along with time called infrared signature, which reflects infrared intensity. For better understanding of infrared signature, research on parameters which affect the infrared intensity are of great importance. In this paper, an infrared radiation intensity model is introduced to describe the intensity change with projection area, observing angle, temperature and so on. Take the cone object as an example for simplicity. The surface of object is divided into hundreds of patches in the calculation of projection area. Transition of radar coordinate system, the object's local coordinate system and the reference coordinate system is involved in the determination of observing angle. These parameters are assumed according to relevant literature, and the simulation of the infrared intensity fluctuation is conducted. From this model, estimation of emissivity-area product and temperature is figured out using maximum likelihood estimation with intensity series. To decrease the degree of sensitivity to noise, Tikhonov-regularized maximum likelihood estimation is introduced. In order to separate temperature and emissivity-area product, observations are made simultaneously in two wavebands. Experiments have been conducted to confirm the effectiveness of estimation algorithm presented. Estimation performance is measured by RMSE (root mean square error). Simulation results show that as the SNR increases, the RMSE decreases, illustrating more and more accurate estimation results. As future research, the motion and shape parameters of exo-atmosphere objects can be particularly studied which lead to further understanding of the infrared signature.