In image deblurring problems, both local and nonlocal regularization priors are well studied. Local regularization prior assumes piecewise smoothness and transform-based sparsity, while the nonlocal one exploits self-similarity of images. We proposed a mixed regularization model which incorporates the advantages of both local adaptive sparsity prior and nonlocal sparsity prior resulting from the nonlocal self-similarity, and thus encourages a solution to simultaneously express both the local and nonlocal natures of images. The deblurring problem with mixed regularization can be transformed into a constrained optimization problem with separable structure via the variable splitting. Then this constrained optimization problem is solved by the alternating direction method of multipliers. Experimental results with a set of images under varying conditions demonstrate that the proposed method achieves the state-of-the-art deblurring performance.
Geosynchronous satellite has obvious limitations for the weight and the scale of payloads, and large aperture optical system is not permitted. The optical diffraction limit of small aperture optical system has an adverse impact on the resolution of the acquired images. Therefore, how to get high resolution images using super-resolution technique with the acquired low resolution images becomes a popular problem investigated by researchers. Here, we present a novel scheme to acquire low resolution images and process them to achieve a high resolution image. Firstly, to acquire low resolution images, we adopt a special arrangement pattern of four CCD staggered arrays on the focal plane in the remote sensing satellite framework .These four CCD linear arrays are parallelized with a 0.25√2 pixel shift along the CCD direction and a 1.25 pixel shift along the scanning direction. The rotation angle between the two directions is 45 degree. The tilting sampling mode and the special arrangement pattern allow the sensor to acquire images with a smaller sampling interval which can give the resolution a greater enhancement. Secondly, to reconstruct a high resolution image of pretty good quality with a magnification factor 4, we propose a novel algorithm based on the iterative-interpolation super resolution algorithm (IISR) and the new edge-directed interpolation algorithm (NEDI). The new algorithm makes a critical improvement to NEDI and introduces it into the multi-frame interpolation in IISR. The algorithm can preserve the edges well and requires a relatively small number of low-resolution images to achieve better reconstruction accuracy .In the last part of the paper, we carry out a simulation experiment, and use MSE as the quality measure. The results demonstrate that our new scheme substantially improves the image resolution with both better quantitative quality and visual quality compared with some previous normal methods.
In this paper, we present a novel iterative interpolation super-resolution algorithm based on the edge-directed interpolation algorithm (NEDI) and the iterative-interpolation super-resolution algorithm (IISR). Our proposed algorithm introduces the NEDI which has only been used for single-image interpolation in previous researches into multi-frame interpolation area of the IISR by way of mapping two images with a 0.5 pixel shift along both directions into a high-resolution grid and populating the grid using improved NEDI. The novel algorithm employs an iterative interpolation process which can be divided into two steps. Firstly, we map low resolution images into the high resolution grid, and use the new interpolation method based on improved NEDI to interpolate the grid to create the first approximation image. Secondly, to satisfy the observation constraints provided by the given low-resolution images, we implement the iteration procedure during which the error vector between the simulated low resolution image and the original one is reconstructed into a high-resolution error image using the same interpolation technique as the first approximation image. After several iteration cycles, the reconstructed high resolution image converging to the real scene is achieved. Absorbing the merits of NEDI and the iterative procedure as well as the improvement to them, the proposed algorithm can preserve the edges well and achieve higher reconstruction accuracy without amplifying the noise and with very few artifacts though using insufficient low resolution images. At last, we carry out a simulation experiment with grayscale images and color images, and the new algorithm demonstrates much better performance compared with some previous normal methods, and the application to noise corrupted low resolution images confirms its robustness.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
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.