基于多次收敛分割反卷积算法的大气色散补偿方法

Atmospheric Dispersion Compensation Based on Multiple Convergence Deconvolution Algorithm

摘要:

在地基望远镜对太空目标的宽光谱成像观测过程中,大气色散会严重影响低仰角下图像的信噪比和清晰度。传统的色散棱镜补偿方法存在着生产加工难度大、控制系统复杂、装配困难等问题,不利于移动和小型化的地基望远镜的应用。提出了一种基于图像反卷积的大气色散修正方法,不需要增加额外补偿设备,能够在低信噪比情况下有效修复色散图像。该方法首先利用多次收敛分割算法对图像进行星点分割处理,随后利用图像反卷积的方法,对图像进行色散补偿。在1.8m望远镜系统上对恒星的观测和色散补偿实验结果表明,该方法在不同信噪比下均有复原效果,平均恢复出目标90%的能量,信噪比提高到3倍以上,恢复效果与色散棱镜补偿法相当。

Abstract:

In the wide-spectrum imaging observation of space target by ground-based telescope, the atmospheric dispersion will seriously affect the signal-to-noise ratio (SNR) and sharpness of image under low elevation angles. The traditional dispersion prism compensation method presents such problems as difficult processing, complicated control system and difficult assembly, resulting in a difficult problem for applications of mobile and miniaturized ground-based telescope. In this paper, an atmospheric dispersion correction method based on image deconvolution is proposed, which can effectively repair the dispersion image at low SNR without adding additional compensation equipment. The stars in the image are first processed by the multi-convergence segmentation algorithm, and then the image is compensated by the deconvolution method. Experimental results of star observation and dispersion compensation on the 1.8m telescope system show that the method has the ability to correct the atmospheric dispersion effect under different SNR, the average energy of targets are restored to more than 90%, and the SNR increases above three times. The compensation effect is equivalent to the dispersion prism compensation method.