Motion Estimate Based on CT Reconstruction in the Region of Interest
-
摘要: 为了对感兴趣区域运动信息进行估计,首先扩展中心切片定理并推导微分约束条件,然后用有限Hilbert变换算法求出方位角与径向图像导函数,最后用改进的光流方程求解极坐标系下的运动场。与一般运动估计算法相比,该算法不使用图像域差分运算计算导函数,因此具有更高精度,并且极坐标系下的运动估计更加适合医学诊断需要。数值仿真也验证了该算法的可行性。Abstract: To estimate the motion information in the region of interest(ROI),firstly central slice theorem is generalized and differential constraint condition is derived. Secondly,the finite Hilbert transform algorithm is adopted to evaluate the partial derivatives of image function along the azimuthal and radial directions,respectively. Finally,the motion vector feld in the polar coordinate system is determined by an improved equation of opticN flow. Compared wim general motion estimate methods,the proposed method is not required to calculate the derivatives using the difference operation in image domain,therefore may show a much higher accuracy,and the motion vector field in the polar coordinate system is much suitable for medical diagnostic requirement. The feasibility of the proposed algorithm has been also validated by using numerical simulation
-
Keywords:
- central slice theorem /
- Hilbert transform /
- motion estimate /
- optical flow
-
计量
- 文章访问数: 720
- HTML全文浏览量: 4
- PDF下载量: 5
下载:
