Half-blind remote sensing image restoration with partly unknown degradation

Meihua Xie; Fengxia Yan

doi:10.1117/12.2265349

23 January 2017 Half-blind remote sensing image restoration with partly unknown degradation

Meihua Xie, Fengxia Yan

Proceedings Volume 10322, Seventh International Conference on Electronics and Information Engineering; 1032213 (2017) https://doi.org/10.1117/12.2265349
Event: Seventh International Conference on Electronics and Information Engineering, 2016, Nanjing, China

Abstract

The problem of image restoration has been extensively studied for its practical importance and theoretical interest. This paper mainly discusses the problem of image restoration with partly unknown kernel. In this model, the degraded kernel function is known but its parameters are unknown. With this model, we should estimate the parameters in Gaussian kernel and the real image simultaneity. For this new problem, a total variation restoration model is put out and an intersect direction iteration algorithm is designed. Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index Measurement (SSIM) are used to measure the performance of the method. Numerical results show that we can estimate the parameters in kernel accurately, and the new method has both much higher PSNR and much higher SSIM than the expectation maximization (EM) method in many cases. In addition, the accuracy of estimation is not sensitive to noise. Furthermore, even though the support of the kernel is unknown, we can also use this method to get accurate estimation.

Citation Download Citation

Meihua Xie and Fengxia Yan "Half-blind remote sensing image restoration with partly unknown degradation", Proc. SPIE 10322, Seventh International Conference on Electronics and Information Engineering, 1032213 (23 January 2017); https://doi.org/10.1117/12.2265349

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available