一种新的轮廓提取方法及其在MR-CT配准中的应用

提出一种新的轮廓提取算法,并将这种算法应用到刚体配准.这种新的轮廓提取算法通过属性的大小自动获得属性算法中的属性阚值及其对应的灰度阚值,对灰度阙值对应的层集进行属性运算后再应用梯度算子得到轮廓.该算法具有强抗噪性而且轮廓边缘保持完好.本文还提出该算法的性质并证明.这种算法提取的脑MR-CT图像的轮廓非常相似,即将多模态配准转化为单模态配准.实验证明配准精度大大提高.

New Contour Extracting Method and its Application in Registration of MR-CT

Presented a new contour extraction method to get feature for rigid registration.The new contour extraction method gets the attribute threshold and the gray threshold by the sizes of the attribute automatically.The level set corresponding the gray threshold is operated by the attribute operation,then the gradient operator is used.The method has the strong anti-noise property and the shapes of contours are kept good.Some important properties are proposed and proved.The contours extracted by our method are ver...