A signal model for the IR signature of exo-atmosphere micro-motion targets

Discriminating exo-atmosphere micro-motion targets is a key technology for precise guidance systems, infrared surveillance systems, and satellite remote sensing systems. Commonly, static IR images are insufficient to discriminate different targets in the field of view. The infrared signature, i.e., the time sequence of the target point's gray level in the infrared images serve as potential discriminants and has been used in many systems. To discriminate the object based on the infrared signature, it is necessary to characterize how the object will appear to the given infrared sensor as a function of engagement geometry, object dynamics, object properties and observation period. Based on the analysis of the infrared sensing process, a signal model is proposed in this paper to approximate the generative process of the infrared signature. The model is based on three observations. First, when the thermal conductivity of the object's surface material is sufficiently high (e.g., metal), the temperatures at different positions of the surface approach to be the same. Second, commonly, IR signature of several seconds long is used to recognize the object. During this short period, the temperatures will not change two much. Third, due to the long observing distance (e.g., 100 km), the direction from target to sensor changes very little during the observing interval. The model shows that the radiation received by the sensor is proportional to projection area of the target during the observing period. With respect to micro-motion, the projection area is a function of shape and motion parameters. So the radiation is represented as a function of the motion and shape parameters. In the experiments, the signature generated by the model and the simulated signature based on finite element analysis are compared, which verifies the reasonable of our model. More importantly, the model may be potential of estimating the shape and motion parameters.