Abstract:
This paper presents a new image registration method based on minimal probability distance and improved Partial Volume(PV) interpolation. Powell optimization algorithm is used to search the symmetric Kullback Leibler(KL) distance to get the minimal probability distance, which is used to be the criteria of registration. An adaptive PV interpolation method is given to improve the robustness of registration. Experimental result shows that this method can effectively reduce the time cost and improve the correctness of registration compared to the method based on mutual information.
Key words:
Mutual Information(MI),
image registration,
minimal probability distance,
Partial Volume(PV) interpolation
,
Powell optimization
摘要： 提出一种基于最小概率距离和改进部分体积(PV)插值的图像配准方法。采用Powell优化算法迭代搜索对称KL距离的最小值，获取最小概率距离，将其作为配准测度，并利用改进的PV插值算法提高图像配准的鲁棒性。实验结果表明，与基于互信息的图像配准方法相比，该方法能有效地减少耗费时间，提高配准精度。
关键词:
互信息,
图像配准,
最小概率距离,
部分体积插值,
Powell优化
CLC Number:
WANG He-Chao, WU Shuan-Hu, ZHANG Huo-Xiang. Image Registration Based on Minimal Probability Distance and Improved PV Interpolation[J]. Computer Engineering, 2011, 37(21): 188-190.
王鹤涛, 武栓虎, 张越翔. 基于最小概率距离和改进PV插值的图像配准[J]. 计算机工程, 2011, 37(21): 188-190.