|
Using the observation data of various detectors to identify reentry vehicles, heavy and light decoys, and separate debris is a key task in space situational awareness. During the flight, the space targets are always in a rotating or rolling state (called micromotion). micromotion can reflect the physical attribute information such as mass distribution and shape of different targets, which provides important essential characteristics for identifying space targets. Infrared sensor has the advantages of working all day, long detection distance, and small load. The image data obtained by it can be used to estimate the temperature, radiation, and other information, but the research on estimating the target micromotion characteristics from the multi-infrared images is rarely mentioned. Therefore, aiming to solve the problem of micromotion period estimation of space infrared moving targets under long-distance observation, firstly, considering the factors such as flight scene, target shape and micromotion, the infrared radiation and imaging models of space moving targets under micromotion state are established according to the micromotion dynamics, temperature and imaging relationship; Secondly, the period of infrared radiation extracted from multi-frame images is estimated. Through theoretical analysis, it is pointed that the assumption that there must be a similarity between the sample sets sampled by period length is the main reason for the doubling misjudgment of the average amplitude difference (AMDF) function, and there is also a false valley misjudgment problem in AMDF. The cyclic average amplitude difference function (CAMDF) is used to estimate the micromotion period of multi-shape objects, which can not only effectively decrease the double misjudgment of the period but also solve the misjudgment of false valley estimation points. Finally, a semi-physical simulation platform for space infrared dim moving target detection and recognition is designed and built, and the experimental data is used to verify the effectiveness of CAMDF in estimating the micromotion period. The results show that when the signal-to-noise ratio(SNR) of the simulated infrared radiation is greater than 15, the average accuracy of CAMDF is greater than 90%; Experimental data of five shape objects is used to verify the algorithm, and the average relative error is about 6%. It shows that the algorithm can better estimate the micromotion period of space targets.
|
