The midcourse ballistic closely spaced objects（CSO） create blur pixel-cluster on the space-based infrared focal plane,making the super-resolution of CSO quite necessary.A novel algorithm of CSO joint super-resolutio...The midcourse ballistic closely spaced objects（CSO） create blur pixel-cluster on the space-based infrared focal plane,making the super-resolution of CSO quite necessary.A novel algorithm of CSO joint super-resolution and trajectory estimation is presented.The algorithm combines the focal plane CSO dynamics and radiation models,proposes a novel least square objective function from the space and time information,where CSO radiant intensity is excluded and initial dynamics（position and velocity） are chosen as the model parameters.Subsequently,the quantum-behaved particle swarm optimization（QPSO） is adopted to optimize the objective function to estimate model parameters,and then CSO focal plane trajectories and radiant intensities are computed.Meanwhile,the estimated CSO focal plane trajectories from multiple space-based infrared focal planes are associated and filtered to estimate the CSO stereo ballistic trajectories.Finally,the performance（CSO estimation precision of the focal plane coordinates,radiant intensities,and stereo ballistic trajectories,together with the computation load） of the algorithm is tested,and the results show that the algorithm is effective and feasible.展开更多
基金supported by China Postdoctoral Science Foundation(20080149320080430223)the Natural Science Foundation of An-hui Province (090412043)
文摘The midcourse ballistic closely spaced objects（CSO） create blur pixel-cluster on the space-based infrared focal plane,making the super-resolution of CSO quite necessary.A novel algorithm of CSO joint super-resolution and trajectory estimation is presented.The algorithm combines the focal plane CSO dynamics and radiation models,proposes a novel least square objective function from the space and time information,where CSO radiant intensity is excluded and initial dynamics（position and velocity） are chosen as the model parameters.Subsequently,the quantum-behaved particle swarm optimization（QPSO） is adopted to optimize the objective function to estimate model parameters,and then CSO focal plane trajectories and radiant intensities are computed.Meanwhile,the estimated CSO focal plane trajectories from multiple space-based infrared focal planes are associated and filtered to estimate the CSO stereo ballistic trajectories.Finally,the performance（CSO estimation precision of the focal plane coordinates,radiant intensities,and stereo ballistic trajectories,together with the computation load） of the algorithm is tested,and the results show that the algorithm is effective and feasible.
基金The General Financial Grant from China Postdoctoral Science Foundation(No.2013M532167)the Hunan Provincial Innovation Foundation for Postgraduates(No.CX2013B019)the Fund of Innovation,NUDT(No.B130403)
文摘利用红外成像系统跟踪多个相距很近的点目标时,目标在成像面上的弥散像会发生交叠,导致成像系统无法有效分辨这些目标。本文提出了一种分辨这类小间距目标(Closely Spaced Objects,CSO)的新方法,通过建立小间距目标的成像模型,采用Gibbs抽样方法对小间距目标在焦平面上的中心位置和响应幅度进行估计,并利用贝叶斯信息准则(Bayesian Information Criterion,BIC)检测目标数目。针对仿真生成的红外图像进行了仿真实验,实验结果表明本文方法对小间距目标的分辨是有效的。
