摘要
针对单一特征引导医学图像配准的准确度有限性,提出一种使用多种特征的图像配准方法.特征提取采用半自动化方式,使操作者能够精确地获得图像特征.对提取的曲线对采用非均匀三次B样条建立模型,使曲线对具有相同的参数区间;非均匀的曲线离散机制保证了离散后的点集在尽可能忠实于原曲线,同时又满足图像配准的要求;通过不断地改善曲线间差异最大的区域并持续添加的约束条件,迭代地求解变换函数.实验结果表明:该算法结合了基于特征点与曲线2种配准算法的优点,既保证了基于点配准的精确性,又兼有基于曲线配准的鲁棒性,是一种有效的医学图像配准方法.
To overcome the limitation of medical image registration introduced by a single kind of feature, an image registration method using multi-features is proposed in the paper. In the method, the operator exactly extracts the features from the images with semi-automatic extraction method. The extracted curve- pairs are modeled by non-uniform cubic B-splines so that they have the same parameter space. The mechanism of non-uniform subdivision of curves ensures that the discrete points can match the original curve as close as possible and fulfill the requirement of image registration. The transformation function is solved iteratively by continuously improving the regions of maximum difference between two curves and adding constraint conditions. Experimental results show that, by taking the advantages in the registration of the accuracy based-on points and the robustness based-on curves, our algorithm is an effective medical image registration algorithm.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2007年第9期1126-1131,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
国家"九七三"重点基础研究发展规划项目(2006CB303106)
浙江省自然科学基金(M603129).
关键词
医学图像配准
特征点
曲线
薄板样条
medical image registration, feature points, curves, thin plate splines