Unified Functional Framework for Restoration of Image Sequences Degraded by Atmospheric Turbulence
We propose a unified functional to address the restoration of turbulence-degraded images. This functional quantifies the association between a given image sequence and a candidate latent image restoration. Minimizing the functional using the alternating direction method of multipliers (ADMM) and Moreau proximity mapping leads to a general algorithmic flow. We show that various known algorithms can be derived as special cases of the general approach. Furthermore, we show that building-blocks used in turbulence recovery algorithms, such as optical flow estimation and blind deblurring, are called for by the general model. The main contribution of this work is the establishment of a unified theoretical framework for the restoration of turbulence-degraded images. It leads to novel turbulence recovery algorithms as well as to better understanding of known ones.
This research was supported in part by the Blavatnik Interdisciplinary Cyber Research Center, Tel Aviv University.
- 2.Aubailly, M., Vorontsov, M.A., Carhart, G.W., Valley, M.T.: Automated video enhancement from a stream of atmospherically-distorted images: the lucky-region fusion approach. In: Proceedings of the SPIE, vol. 7463 (2009)Google Scholar
- 11.Gadot, D., Wolf, L.: Patchbatch: a batch augmented loss for optical flow. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)Google Scholar
- 13.Hirsch, M., Sra, S., Scholkopf, B., Harmeling, S.: Efficient filter flow for space-variant multiframe blind deconvolution. In: Computer Vision and Pattern Recognition (CVPR), pp. 607–614, June 2010Google Scholar
- 15.Joshi, N., Cohen, M.: Seeing Mt. Rainier: lucky imaging for multi-image denoising, sharpening, and haze removal. In: Proceedings of the IEEE ICCP (2010)Google Scholar
- 17.Mao, Y., Gilles, J.: Turbulence stabilization. Proc. SPIE 8355, 83550H–83550H-7 (2012)Google Scholar
- 23.Zak, N.: Restoring an image of a moving object from a turbulence-distorted video. Master’s thesis, School of Electrical Engineering, Tel Aviv University, Israel (2015)Google Scholar