Dim Small Moving Target Detection In Complicated Sky Background

Dim moving target detection is widely used in aerospace and military fields,for national security and strategic balance of space to the moon and the urgent need of wide-area monitoring for space extreme distance to the moon strong background,the weak target in the target detection and recognition and high noise caused by the polar SNR moving target is difficult to detect problems,carried out under the background of complex star field of faint space small moving target detection research,analyzes the background and the characteristics of faint moving targets and construct the interaction model of target and the background,implementation under the complex background of sky is extremely low SNR moving target detection,finally achieve the purpose of improve the level of wide-area space target detection and recognition.The method presented in this paper is mainly to analyze and process the signal in time domain and to use the disturbance brought by the moving target to the signal in the time interval when it enters and leaves the pixel to reflect the existence of the target.This paper puts forward a kind of dim small moving target based on the analysis of the fluctuation in a waning space correlation detection algorithm,carried out based on the goal of light field fluctuation signal and background noise in the time domain characteristics analysis,build the timing signal correlation statistical model,through the way of dividing time window to extract the local fluctuation characteristics,and realize the faint small moving target detection.The experimental results show that the detection rate and false alarm rate of this method are significantly better than those of other classical methods.It has a good detection effect in both simulation data and real data.The detection rate and false alarm rate of this method are 87.8% and 28.84% in semiphysical simulation experiment.In the real asteroid data,the detection rate is 70.56%,and the false alarm rate is 3.62%.We finally propose a small moving target detection algorithm based on Hidden semi-Markov Model(Hidden semi-Markov Model,HSMM).According to HSMM,the interaction model of the target signal and the background noise in the time series is constructed,and then the time series signal in the pixel is modeled and the state is divided.A series of characteristics such as the fluctuation of the time sequence signal in the window are used as the criterion for state distinction during the detection period.The Baum-Welch algorithm is used to solve the most likely hidden state sequence corresponding to the observed time sequence,and then the state where the target exists is analyzed.The experimental results show that the detection rate of this method in the near-Earth asteroid detection experiment reaches 80.35%,and the false alarm rate is1.38%,which verifies the effectiveness of the method.