论文部分内容阅读
针对低信杂比漂移扫描星图中小目标检测易受恒星星像干扰的问题,提出一种基于单帧候选检测及多帧跟踪判决的实时处理方法。分析星图难以多帧匹配去恒星星像的原因,采用单帧处理的方式,通过构造星像端点掩膜滤除单帧检测中的虚假目标。将单帧得到的候选目标进行粒子滤波的粒子生成,运用目标运动信息构造似然函数分布图更新粒子权重,通过粒子滤波实现了漂移扫描星图中的小目标检测。实际低信杂比星图的实验结果和与相关方法的比较表明,所提方法有效提高了弱小目标的检测能力。
Aiming at the problem that the detection of small target in the low signal-to-noise ratio (SAR) scan map is easily disturbed by the star image, a real-time processing method based on single-frame candidate detection and multi-frame tracking decision is proposed. This paper analyzes the reason why the star chart is difficult to match the star images of the multi-frame matching, and uses the single-frame processing method to filter out the false targets in the single-frame detection by constructing the star endpoint mask. Particle filter is used to generate particle filter for the candidate target obtained from a single frame. The particle weight is updated by constructing the distribution of likelihood function using the target motion information, and the small target detection in the drift scan star map is achieved by particle filter. The experimental results of the actual low signal-to-noise ratio chart and the comparison with the related methods show that the proposed method can effectively improve the detection ability of the weak target.