An improved curvelet thresholding denoising algorithm for astronomical image

In astronomical image processing, to solve the problem of slow convergence speed and poor denoising performance in compressed sensing iterative curvelet thresholding (ICT) algorithm, an improved ICT reconstruction algorithm with high performance is proposed. A Dai-Yuan step size is used by algorithm to accelerate its convergence speed. To improve the quality of the reconstructed image, a new curvelet threshold is proposed to select the curvelet coefficients of astronomical image. Meanwhile, the total variation method is employed to adjust, the reconstructed image in each iteration for suppressing the pseudo-gibbs effect in the reconstructed image. Number experimental results demonstrate that, the algorithm proposed is superior to the traditional ICT algorithm, which can achieve better denoising performance with a fast, convergence speed and effectively protect the image detailed features. Furthermore, even with a lower compression ratio, the proposed algorithm can still obtain a higher peak signal to noise ratio (PSNR).