Abstract: A new image denoising method based on Contour Bougie (CB) elements morphology was proposed. Considering the image characteristics on different scales, the morphological multi-scale algorithm based on contour structure elements was employed to remove the noise in image, and then the weighted fusion of the denoised image on different scales was done. A large number of experiments show that the proposed method can not only suppress many kinds of image noise but also protect the image detail effectively.

Key words: mathematical morphology, muli-scale, image denoising, filtering

摘要： 提出一种新的基于轮廓结构(CB)元素形态学的多尺度图像降噪方法。考虑到图像在不同尺度下的特性，用基于轮廓结构元素形态学的多尺度算法滤除图像噪声，然后将不同尺度下的降噪图像加权融合在一起，从而得到最终的降噪图像。通过大量实验证明该方法能够有效去除多种不同类型的图像噪声，并在抑制噪声的同时很好地保护了图像的细节和边缘信息。

关键词: 数学形态学, 多尺度, 图像降噪, 滤波