A detection algorithm for faint sources based on 1-d projection for a lobster-eye x-ray imaging system
A detection algorithm for faint sources based on 1-d projection for a lobster-eye x-ray imaging system
复制标题
龙虾眼 X 射线成像系统中基于一维投影的微弱源检测算法
DOI：
10.1117/12.2560604
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Serino Motoko
中科院分区：
文献类型：
--
作者：
Sawano Tatsuya;Yonetoku Daisuke;Arimoto Makoto;Li Jin;Mihara Tatehiro;Ogino Naoki;Sakamoto Takanori;Serino Motoko
本文提出了一种优化的基于龙虾眼光学的x射线宽视场成像探测器源检测算法，以实现更高的灵敏度。在我们的方法中，我们采用感兴趣区域（ROI）的两个部分来测试超过一定阈值水平的x射线事件的数量。由于我们构成了ROI的两个部分都需要光子过量的条件，因此我们可以降低检测阈值水平并降低误报率。我们采用了两种以ROI为一部分的比较方法，并对它们的性能进行了比较。我们为两种可比较的检测方法以及我们提出的方法制定了适当的阈值水平和灵敏度。我们发现，在提议的高高达任务的名义情况下，我们的方法的检测灵敏度比可比方法最多提高了约30%。我们还发现，哪种检测方法具有更好的灵敏度取决于背景事件率。我们通过与蒙特卡罗模拟的HiZ-GUNDAM条件进行比较，验证了该公式的有效性。该公式可应用于未来任何使用LE光学的任务，以设计适合优化灵敏度的检测算法。
We propose an optimized source detection algorithm with an X-ray wide-field imaging detector based on lobstereye (LE) optics to realize better sensitivity. In our method, we take two parts of region of interest (ROI) in which we test the number of X-ray events exceed a certain threshold level. Since we compose the condition that the excesses of the photons are required for the both parts of the ROI, we can lower the detection threshold level with a less false alert rate. We take two comparative methods in which the ROI consists of one part, and compared the performance of them. We formulated an appropriate threshold level and sensitivity for two comparable detection methods as well as our proposed method. We found that the detection sensitivity of our method is improved by a factor of about 30% at most than that of the comparable methods in the nominal case of the proposed HiZ-GUNDAM mission. We also found that which detection method has better sensitivity depends on the background event rate. We checked that the formulation works well by comparing to a Monte Carlo simulation in the case of the HiZ-GUNDAM condition. The formula can be applied to any future missions with LE optics to design which detection algorithm is suitable for optimizing sensitivity.