CAJ | 학술논문

提出了一种新的多幅图像配准方法,归一化互信息向量熵方法.这种方法先计算任意两幅图像间的联合概率分布,然后根据联合概率分布计算它们间的归一化互信息,把所有两幅图像组合得到的归一化互信息组成一个向量,最后计算该归一化互信息向量的熵.最大熵对应最佳配准位置.通过对人体脑部图像的刚体配准实验,从函数曲线、计算时间和配准精度方面,对新方法和其它三种方法进行了分析和比较.实验结果表明,新提出的方法可以提高配准精度、减少配准时间.
제출료일충신적다폭도상배준방법,귀일화호신식향량적방법.저충방법선계산임의량폭도상간적연합개솔분포,연후근거연합개솔분포계산타문간적귀일화호신식,파소유량폭도상조합득도적귀일화호신식조성일개향량,최후계산해귀일화호신식향량적적.최대적대응최가배준위치.통과대인체뇌부도상적강체배준실험,종함수곡선、계산시간화배준정도방면,대신방법화기타삼충방법진행료분석화비교.실험결과표명,신제출적방법가이제고배준정도、감소배준시간.
A novel method for multi-image registration is proposed,which is called the entropy of normalized mutual information vector.This method first calculates joint probability distribution of any two images,and then calculates the normalized mutual information according to it.All the normalized mutual information of two images forms a vector,the normalized mutual information vector.At last the entropy of that vector is calculated.The maximal entropy corresponds to the optimal registration solution.The function curves,computing time and registration accuracy are studied by applying the new method and other three methods to rigid registration of brain images.The obtained results show that the proposed method can improve registration accuracy and decrease registration time.