提出一种基于图像特征的一致点漂移匹配方法进行3维形状点集对齐,在匹配过程中结合点的几何空间信息和图像特征信息构造目标函数,依据点之间的图像特征差异调整原始一致点漂移匹配方法中的高斯混合模型。使用3维前列腺和肝脏点集进行匹配的仿真实验,结果表明本文方法可有效减少匹配误差,其中肝脏点集匹配误差从1.84 mm降低到1.54 mm,前列腺点集匹配误差从0.83 mm降低到0.60 mm。利用本文提出的形状点集对齐方法建立活动外观模型,对3维前列腺CT图像进行分割,分割精度有一定提高,即体素正确覆盖率从88.7%提高到90.2%。

A key step of constructing an active appearance model is acquiring a set of appropriate training shapes with well-defined correspondences. In this paper, we introduce a new point correspondence method (FB-CPD), which can improve the accuracy of the coherent point drift (CPD) by using the image feature information. The objective function of the proposed method is defined by both of the geometric spatial information and the image feature information. The original Gaussian mixture model in the CPD is modified according to the image feature of the points. FB-CPD is tested on the three-dimensioral prostate and liver point sets through the simulation experiments. The registration error can be reduced efficiently by FB-CPD. Moreover, the active appearance model constructed by FB-CPD can obtain fine segmentation, in three-dimensioral CT prostate images. Compared with the original CPD, the overlap ratio of voxels was improved from 88.7% to 90.2% by FB-CPD.