？abry—Perot测风干涉仪数据处理

摘要

Fabry-Perot Interferometer （FPI） is one of the best detectors to explore the middle and upper atmosphere wind. A spaceborne FPI prototype with a large diameter telescope was designed and constructed and the image of interference fringe was acquired. This paper introduces the princi- ple of FPI and a feasible inversion algorithm of the wind velocity in the middle and upper atmosphere. A technique for F-P interference fringe auto-processing is developed. The technique includes filtering, segmentation, thinning, and extracting of interferogram characteristic data. Filtering is composed of median filtering and background compensation. Binary fringe is obtained through selecting the area mean gray scale as threshold for segmentation. Mathematical morphology is fitting for restoring rupture and burr. As thinning is essential for the accuracy of inverting middle and upper atmosphere wind, thinning is discussed in details. Skeletonization thinning is put into practice. At last, characteristic data such as center of circle and radius are extracted. Visual C＋＋ is adopted to develop the FPI interference fringe image process software, and the technique mentioned above has good performance in experimental practice.%针对自主研制的星载法布里一珀罗干涉仪（Fabry—Perot Interferometer，FPI）样机，介绍了Fabry—Perot（F—P）的测风原理和风速反演算法，阐述了F—P测风干涉数据的自动化处理方法，包括干涉条纹图像滤波、分割、细化和特征数据的提取．滤波包括中值滤波和背景补偿，分割采用区域平均灰度值阈值法，借助数学形态学计算填充和消除二值化条纹的断裂点和毛刺．最后，按照骨架提取思想细化成干涉条纹线条，提取其特征数据．应用VC＋＋编程语言开发软件平台，对上述F—P测风仪数据自动化处理方法进行了验证，结果表明其可行性高、适应性强．