基于盲退卷积的AIA图像增强方法  

AIA Images Enhancement Method Based on Blind Deconvolution

作　　者：沐芳 黄欢[1] 尚振宏[1,3] 强振平 Mu Fang;Huang Huan;Shang Zhenhong;Qiang Zhenping(Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China;College of Big Data and Intelligent Engineering, Southwest Forestry University, Kunming 650224, China;Yunnan Key Laboratory of Artifical Intelligence, Kunming University of Science and Technology, Kunming 650500, China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650500 [2]西南林业大学大数据与智能工程学院,云南昆明650224 [3]昆明理工大学云南省人工智能重点实验室,云南昆明650500

出　　处：《天文研究与技术》2022年第1期41-53,共13页Astronomical Research & Technology

基　　金：国家自然科学基金(12063002,11873027,12163004);云南省重大科技专项计划项目(202002AD080001)资助.

摘　　要：太阳图像中存在各种不同尺度、亮度和结构的物理活动现象,由于太阳日冕高动态活动和传感器设备等因素的影响,太阳图像成像质量不佳。根据太阳动力学天文台(Solar Dynamic Observatory,SDO)的大气成像仪(Atmospheric Imaging Assenbly,AIA)拍摄不同波段数据结构的动态范围大、噪声大、结构相对模糊等特点,提出一种基于盲退卷积的图像增强方法。首先对图像进行去噪和降低动态范围的处理,基于图像功率谱的分布假设,从原图中估计点扩散函数(Point Spread Function,PSF)的功率谱;然后使用相位提取算法恢复点扩散函数的相位,再退卷积得出较高质量的目标图像;最后通过轮廓切片分析、功率谱分析以及点扩散函数分析对增强结果进行定量和定性评价。实验结果表明,相比现有的图像增强方法,该方法在有效增强太阳日冕图像细节结构的同时,能够复原原图中因模糊无法识别的结构。There are various physical activities in the solar images,such as different scales,brightness and structure.Due to the high dynamic activity of the solar corona and sensor equipment,the image quality of solar imaging is poor.According to the characteristics of large dynamic range,large noise and relatively fuzzy structure of different waveband data structures captured by the atmospheric imaging asset(AIA)of the solar dynamic Observatory(SDO),an effective SDO/AIA image enhancement method based on blind deconvolution is proposed in this paper.Firstly,the SDO/AIA image is denoised and the dynamic range is reduced.Based on the distribution assumption of image power spectrum,the point spread function(PSF)is estimated from the original image.Then,phase retrieval algorithm is used to recover the phase of PSF,and then deconvolute the high-quality target image;finally,the enhancement results are analyzed quantitatively and qualitatively through slice analysis,power spectrum analysis and PSF analysis.The experimental results show that,compared with the existing image enhancement methods,the proposed method can effectively enhance the detail structure of the solar corona image,and restore the structure that cannot be identified due to the blur of the original image.

关 键 词：太阳图像 大气成像仪 功率谱 盲退卷积 相位提取 

分 类 号：P182.62[天文地球—天文学]