基于自适应字典学习降噪改进的脑MRI图像分水岭精确分割算法研究

发布时间：2018-06-14 10:01

本文选题：字典学习 + 数学形态学　；参考：《激光杂志》2015年01期

【摘要】：图像分割技术是图像处理领域非常活跃的研究课题.但目前还不完善尤其MRI高噪声图像分割还没有给出较好的分割算法,如:脑MRI图像的分割等等.分水岭算法和图像降噪算法在图像分割中有广泛的应用.本文将这两种方法结合起来,并用于脑MRI图像分割,取得较好的分割结果.本文详细论述了字典学习降噪算法的原理,提出一种字典学习降噪和分水岭算法相结合的脑MRI医学图像分割算法.采用字典学习降低原始图像噪声,然后利用形态学算法对降噪后的图像进行形态学处理,通过形态学知识提取图像边界.利用图像的几何特征,去除非目标区域,再采用分水岭变换进行图像分割,并通过脑MRI图像验证了此方法的优势.实验结果进一步验证了其可行性.
[Abstract]:Image segmentation is a very active research topic in the field of image processing. However, at present, there are not perfect segmentation algorithms, such as brain MRI image segmentation and so on. Watershed algorithm and image denoising algorithm are widely used in image segmentation. In this paper, the two methods are combined and applied to brain MRI image segmentation, and good segmentation results are obtained. In this paper, the principle of dictionary learning denoising algorithm is discussed in detail, and a brain MRI medical image segmentation algorithm combining dictionary learning de-noising and watershed algorithm is proposed. The dictionary learning is used to reduce the noise of the original image, and then the morphological algorithm is used to deal with the denoised image, and the edge of the image is extracted by morphological knowledge. Using the geometric features of the image, except the target region, the watershed transform is used to segment the image, and the advantages of this method are verified by the brain MRI image. The experimental results further verify its feasibility.
【作者单位】：成都理工大学工程技术学院;
【基金】：四川省教育厅,项目编号:14ZB0355 乐山市科技局重点项目,项目编号13GZD039 校青年科研基金,项目编号:C122014023等
【分类号】：R445.2

【参考文献】