Improved Algorithm of Adaptive DR Image Denoising Based on Fast Curvelet Transform and Anisotropic Median-diffusion Filtering

摘要

In the direct digital X-ray (DR) imaging process, there are wide variety of complex noises which are due to various causes, such as a CCD camera, the imaging screen, X-ray scatter, control circuit and so on. Therefore, this paper, according to characteristics of the CCD/DR image noise, and combined with adaptive threshold method, research a improved algorithm of adaptive DR image denoising based on fast curvelet transform and anisotropic median-diffusion filtering. This algorithm is simple, direct, more comprehensive and targeted. It is well preserved and enhanced DR image detail and edge information while better denoising to prepare for the post-processing.