In order to improve the performance of the current restoration method for block compressed sensing under the low complexity, we propose a novel variation-based block compressed sensing restoration method. The method decomposes the image into several non-overlapping blocks first, followed by the scanning according to the column and measurement by blocks, respectively, so as to obtain several column vectors of measurement value. The decoding end integrated the column vectors of measurement value received into matrixes, making the sparsity of image regular terms as the prior knowledge and minimizing the augmented Lagrange function as the goal. In this way, the sub-problems can be orderly solved with the variant alternating direction multiplier method, while the column vector space of image blocks was reconstructed. Finally, an anti-scanning was performed before it was combined into images. The innovation point is to apply the total variation model into the restoration framework for block compressed sensing with a small amount of calculation cost, and to extent it to a mixed variation model which can contain multiple regular terms and generality. Contrast to the current restoration algorithm relating to block compressed sensing, the simulation results show that the proposed method can achieve a better SSIM and the fastest restoration speed, while the SSIM and PSNR of the proposed method can achieve the best result.
In order to more accurately realize fusion of remote sensing images, we propose a novel remote sensing images fusion algorithm combining extended non-subsampled shearlet transform (NSST) and modified pulse-coupled neural network (PCNN). Firstly, it makes histogram matching and intensity smoothing and filtering treatment on intensity component and full-color image of multi-spectral image. Secondly, such intensity component and full-color image are decomposed by extended NSST to get corresponding high-frequency and low-frequency coefficients. For low-frequency coefficients, fusion is made by sparse representation; for high-frequency coefficients, a modified pulse-coupled neural network (PCNN) strategy is put forward to process. Finally, the processed result is drawn by inverse transformation of the extended NSST and intensity-hue-saturation inverse transformation. The experimental results show that the proposed algorithm reserves as much spectral information as possible and improve spatial resolution; its visual effects and objective indexes are better than other classical fusion algorithms.
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.