一种基于多尺度几何分析下的图像去噪算法研究

图像的边缘和轮廓包含图像的重要特征信息，当图像中数据混有噪声的时候，对边缘的检测比较困难。在多尺度几何分析理论的基础上，提出了一种基于尺度因子与Contourlet变换的图像去噪算法。实验表明，该算法对高斯白噪声污染的图像去噪具有很好的效果，不仅可以提高处理图像的信噪比，图像的视觉效果也明显改善。

Method for Image Denoising Based on Multiscale Geometric Analysis

The edge and outline of image arc very useful in the pattern recognition. But there arc much difficulties if the signals are polluted by noise. We first introduced the multiscale geometric analysis theory, and then proposed a method for image denoising on scale factor and contourlet transform. Experiments show that the performance of the proposed method is obviously superior to other methods both in vision and in SNR.