最优分布变换与微弱点状动目标检测技术

A Dim Moving Point Target Detection Technique Based on Distribution Function Transform Method

研究了微弱点状动目标序列图像观测模型、目标检测模型和最优分布以及最优分布变换。提出了一种基于准优分布变换的微弱点状动目标检测技术。经过变换 ,未知统计分布变为准高斯分布 ,提高了信噪比 ,改造了分布形状 ,获得了有利分布。并将时空三维搜索简化为沿时间轴投影与二维空域搜索 ,实现了一种快速算法。PC仿真结果与国内外相应文献比较 ,在信噪比、检测概率与虚警概率方面具有更好的性能 ,且算法的实时性好。

A new algorithm based on track-before-detect (TBD) technology is proposed for the detection of dim moving point targets with unknown position and velocity in a sequence of digital images collected from a staring sensor. Through transformation, the pixel statistics are utilized a composite image produced by summing up some differential images, then the detection is performed. This algorithm has been applied to various image sequences with terrestrial and celestial no moving clutter backgrounds that usually are supposed to be non-Gaussian processes. Compared with other algorithms, the proposed algorithm is not only computationally simple, but also has a high performance. Theoretical analysis of the algorithm's performance and experimental results are also given in this paper.