欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
传统的反锐化对图像的高频部分进行增强时，易产生伪影现象，且对噪声非常敏感。针对上述问题，提出一种基于多尺度形态学的图像非线性锐化算法。利用形态学独特的特征提取优势，将传统的反锐化掩膜原理扩展到形态学领域。该算法利用多尺度形态学提取图像的细节特征，并根据梯度信息控制多尺度下的细节特征对输出图像的贡献；同时扩大前景与背景的反差，突出隐藏的信息。实验结果表明，该算法较好地增强了图像的信息，同时抑制了噪声的增加，增强后的图像具有很好的视觉效果。
英文摘要：
When traditional unsharp masking enhances the high-frequency component of the image,it is easy to suffer from artifact,and is very sensitive to noise as well.For these problems,we propose a non-linear image sharpening algorithm which is based on multi-scale morphology.Utilising unique advantage of morphology in feature extraction,it extends the conventional unsharp masking principle to morphological field.This algorithm employs multi-scale morphology to extract the detailed features of the image,and controls the contribution of the detailed features in multi-scale on output image according to gradient information;meanwhile,it expands the contrast between the foreground and the background,and highlights the hidden information.Experimental results demonstrate that this algorithm well enhances the information of the image,and suppresses the addition of noise at the same time.Moreover,the enhanced image has quite good visual effect.
同期刊论文项目
同项目期刊论文
A statistical iteration approach with energy minimization to sinogram noise reduction for low-dose X
Regularized maximum likelihood algorithm for PET image reconstruction using a detail and edges prese
Application of regularized maximum likelihood algorithm in PET imagereconstruction combined with non
期刊信息
位置：