非下采样剪切波域的临近支持向量机去噪方法

Image denoising method using PCC classification in nonsubsampled Shearlet domain

非下采样剪切波(nonsubsampled Shearlet)是一种优秀的多尺度几何分析工具,其不仅可以检测到所有奇异点,而且能够自适应跟踪奇异曲线方向。基于非下采样剪切波,提出了一种使用带有一致性的临近支持向量机(a Proximal Classifier with Consistency,PCC)的图像去噪算法。首先,应用非下采样剪切波把含噪图像分解成不同尺度不同方向的子带;其次,非下采样剪切波系数通过PPC训练被分成两类(无噪系数和噪声系数);最后应用自适应阈值对含噪系数进行去噪。仿真实验结果表明,本文算法不仅拥有较强的抑制噪声能力,而且具有较好的边缘保护能力。

Nonsubsampled Shearlet transform is a kind of excellent multiresolution analysis tool, and it can provide nearly optimal approximation for a piecewise smooth function. Based on nonsubsampled Shearlet transform, an image denoising method using Proximal Classifier with Consistency（PCC） is proposed. Firstly, the noisy image is decomposed into different subbands of frequency and orientation responses using the nonsubsampled Shearlet transform.Secondly, the nonsubsampled Shearlet transformdetail coefficients are divided into two classes（edge-related coefficients and noise-related ones） by PCC training model. Finally, the detail subbands of nonsubsampled Shearlet transform coefficients are denoised by using the adaptable Bayesian threshold. Extensive experimental results demonstrate that the proposed method can obtain better performances in terms of both subjective and objective evaluations than those state-of-the-art denoising techniques. Especially, the proposed method can preserve edges very well while removing noise.