CS high-noise astronomical image denoising reconstruction method combined with fractional order total variation

Abstract

The invention provides a CS high-noise astronomical image denoising reconstruction method combining fractional order total variation. The method comprises: initializing reconstruction image parametersand noise-containing observation values according to an original astronomical image; performing multi-scale decomposition on the high-resolution astronomical image under a high-noise condition by utilizing curvelet transform with an adaptive filtering operator to obtain a curvelet coefficient; utilizing an iterative shrinkage threshold algorithm to iteratively update the curvelet coefficient to obtain an image curvelet coefficient; carrying out noise reduction processing on the image curvelet coefficient by using a descending curvelet threshold operator; performing processing by utilizing curvelet inverse transformation to obtain a reconstructed astronomical image; performing feature adjustment on the reconstructed astronomical image by using a fractional order total variation method to obtain an adjusted reconstructed image; and determining an iterative stop condition. According to the method, the denoising capability of the high-resolution astronomical image under the high-noise condition is effectively improved, the denoising reconstruction problem of the astronomical image under the high-noise condition in deep space exploration is solved, and the method is easy to implement and high in robustness.

Description

Technical field [0001] The present invention relates to the technical field of image processing, in particular to a CS high-noise astronomical image denoising reconstruction method combined with fractional total variation, which combines compressed sensing (CS) and total variation to perform denoising processing on astronomical images to achieve high The fast denoising reconstruction of noisy astronomical images, especially related to the fast denoising ability of high-resolution astronomical images under high-noise conditions. Background technique [0002] With the continuous development of space science and technology, deep space exploration has become more and more important. In order to be able to actively develop and utilize space resources, countries all over the world are actively exploring extraterrestrial objects, space resources, etc., and have formulated long-term deep space exploration plans. Astronomical images are of great significance in deep space exploration. Ma...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the technical problem of poor reconstruction quality when existing CS reconstruction methods reconstruct high-noise astronomical images under high-noise conditions, the present invention proposes a CS high-noise astronomical image denoising and reconstruction method combined with fractional total variation. In the stage, an adaptive filter operator is introduced to perform sparse transformation and denoising processing on high-noise astronomical images; then, the fractional total variation method is introduced into the iterative shrinkage threshold algorithm in CS, and in each iteration, the fractional The first-order total variation method is used to adjust the reconstructed astronomical image, which effectively improves the reconstruction quality of the astronomical image, and finally realizes the high-quality reconstruction of the high-resolution astronomical image under the condition of high noise.

Images