【摘要】 为了解决传统DTI图像分割中更细致边缘信息的丢失问题,提出了新的张量形态学梯度参数,并基于张量相似性形态学梯度和各向异性形态学梯度,采用标记的分水岭算法对DTI图像进行分割。通过对人脑胼胝体图像的分割实验表明,利用新参数TMG-l2和TMG-RA能够更加快速、准确地对DTI图像进行细致边缘轮廓的定位和分割,保护了重要分割区域的边缘信息。更多还原

【Abstract】 To solve the problem of losing detailed edge information inherent in traditional DTI image segmentationmethods, some new morphological gradient tensor parameters are proposed. Based on the tensor similaritymorphological gradients and tensor morphological anisotropic gradients, the tag based on watershed algorithm isapplied in DTI image segmentation. The human brain corpus callosum image segmentation experiments show that thisalgorithm with the TMG-l2 and TMG-RA parameters can quickly and accurately locate and segment the outline ofthe image, and the edge information of the important region is preserved.更多还原