Research On TBD Algorithm Of High Orbit Space Debris In Complex Sky Background

With the increasing frequency of human space activities,the number of space debris is increasing,which has reached the level that we have to pay attention to.In this paper,based on the practical needs of ground-based high orbit space debris photoelectric monitoring,the core technology research of tracking algorithm before detection of high orbit space debris in complex sky background is mainly carried out,which provides technical support for the ground-based photoelectric monitoring task of high orbit space debris.This paper is divided into the following three parts:In the first part,the algorithm of suspected target extraction from high orbit space debris in complex sky background is studied.Firstly,an adaptive fast registration algorithm for complex star image is proposed to solve the influence of star drift in groundbased optical telescope image on the extraction of suspected debris target in the target tracking mode.Secondly,an enhanced dilation difference algorithm is proposed to improve the traditional dilation difference algorithm,which can well solve the residual image caused by the time-varying brightness of stars.Therefore,it can improve the suppression effect of complex stellar background.Finally,the algorithm of fragment detection and high-precision centroid location based on inter frame correlation algorithm is designed.In the second part,based on the effective extraction of debris target,the tracking method before high orbit space debris detection is studied.Firstly,the basic principle of the high orbit space debris tracking algorithm based on the traditional data association algorithm is described,and the performance of the algorithm is summarized by combining the measured data,mainly including the classical Global Nearest Neighbor,Multiple Hypothesis Tracking,and Joint Probabilistic Data Association algorithm;secondly,considering the shortcomings of the traditional tracking algorithm in the high orbit space debris tracking application,combined with the operation of high orbit space debris.Based on the dynamic and observation characteristics,the state and observation space model of high orbit space debris are established,the δ-Generalized Labeled Multi-Bernoulli algorithm is proposed to be applied to high orbit space debris tracking,and the prediction and update equations in the tracking process are derived.Finally,the advantages of the δ-Generalized Labeled Multi-Bernoulli algorithm are verified based on the measured data.In the third part,the core problems of the track algorithm optimization before the high orbit space debris detection based on tms320c6678 multi-core DSP are studied.Firstly,based on the decomposition of flat structure elements,the high time-consuming morphological background suppression algorithm is optimized to solve the problem of poor real-time background suppression using large-scale structure elements.Secondly,combined with the track algorithm flow before the detection of high orbit space debris,the core problems in the design of multi-core DSP information processing system are studied,mainly including the algorithm structure design,communication port design,storage space allocation and so on.Efficient multi-core pipelining structure,flexible communication port design and reasonable storage space allocation strategy ensure that the scheme has strong universality and stability.