太阳斑点图像重建是天文观测领域中一个重要的研究问题。通过地基光学望远镜获取的天文图像受大气湍流和大气扰动的影响，会发生严重的模糊或降质。剪切波变换是一种多尺度几何分析方法，它比传统的小波变换更符合人类视觉系统的感知特性，能更有效地表示和捕获图像中的边缘、纹理等几何特征，并能充分利用图像自身的几何特性实现对其更为“稀疏”的表示。该方法用于太阳斑点图像融合时存在高频信息缺失、边缘模糊等问题。本文结合剪切波变换和特征信息检测，对多帧太阳斑点图像进行融合处理。首先通过一种简单的剪切波变换域图像融合方法得到一幅初始的融合图像，然后根据所有斑点图与初始融合图像像素间的局部相似程度来获得每一斑点图的有效特征信息区域检测图，并据此将该斑点图中的所有像素分成有效特征信息、有效特征信息与无效信息之间的过渡以及无效信息，最终据此来指导剪切波变换域各子带系数的融合。实验结果表明，本文方法能较好地恢复高频信息，实现太阳斑点图高分辨率重建。
The reconstruction of solar image is an important research problem in the field of astronomical observation. Astronomical images acquired through ground-based optical telescopes are affected by atmospheric turbulence and atmospheric disturbances, and will be severely blurred or degraded. Shearlet transform is a multi-scale geometric analysis method. It is more in line with the perceptual characteristics of the human visual system than traditional wavelet transform. It can more effectively express and capture geometric features such as edges and textures in the image, and can make full use of the image. Its own geometric characteristics realize its more “sparse” representation. This method has problems such as lack of high-frequency information and blurred edges when it is used in the fusion of solar images. In this paper, combining shearlet transform and feature in-formation detection, multi-frame solar images are fused. First, an initial fusion image is obtained through a simple shearlet transform domain image fusion method, and then the effective feature information area detection map of each image is obtained according to the local similarity between all the images and the initial fusion image pixels, and according to this, all pixels in the image are divided into effective feature information, the transition between effective feature information and invalid information, and invalid information, which ultimately guides the fusion of the coefficients of each sub-band in the shearlet transform domain. The experimental results show that the method in this paper can recover high-frequency information well and achieve high-resolution reconstruction of the solar image.
Liu, S.Q., Shi, M.Z., Zhu, Z.H., et al. (2017) Image Fusion Based on Complex-Shearlet Domain with Guided Filtering. Multidimensional Systems and Signal Processing, 28, 207-224. https://doi.org/10.1007/s11045-015-0343-6
Do Minh, N. and Mar-tin, V. (2005) The Contourlet Transform: An Efficient Directional Multiresolution Image Representation. IEEE Transac-tions on Image Processing, 14, 2091-2106. https://doi.org/10.1109/TIP.2005.859376
Geng, P., Wang, Z.Y., Zhang, Z.G. and Xiao, Z. (2012) Image Fusion by Pulse Couple Neural Network with Shearlet. Optical En-gineering, 51, 7005. https://doi.org/10.1117/1.OE.51.6.067005
[10]
Cheng, S., Miao, Q.G. and Xu, P.F. (2013) A Novel Algorithm of Remote Sensing Image Fusion Based on Shearlets and PCNN. Neurocomputing, 117, 47-53. https://doi.org/10.1016/j.neucom.2012.10.025
[11]
Liu, X., Zhou, Y. and Wang, J.J. (2014) Image Fusion Based on Shearlet Transform and Regional Features. AEU: International Journal of Electronics and Communications, 68, 471-477. https://doi.org/10.1016/j.aeue.2013.12.003
