1.no middle initial
1.On each stage, groups of variables are alternately fixed at the iterations to reduce the NMI to linear matrix inequalities (LMIs).
在每一阶段，每一次迭代过程中，通过交替固定NMI的一个变量，使NMI转变为线性矩阵不等式（LMI）。
2.A new algorithm of infrared and visible light image registration based on normalized mutual information(NMI) and wavelet transform is given.
针对灰度和图像特征存在较大差异的可见光和红外图像配准，给出了一种基于归一化互信息与小波变换相结合的快速图像配准算法。
3.This paper proposes an image retrieval method based on NMI (Normalized Monent of Inertia) invariable feature for binary trade mark images.
文中针对二值商标图像，提出一种基于NMI特征的商标图像检索方法。
4.A two-stage homotopy method is then employed to solve the NMI iteratively .
采用两步同伦法迭代来求解非线性矩阵不等式（NMI）。
5.Research of Binary Trade-mark Image Retrieval Method Based on NMI Invariable Feature
基于NMI不变特征的二值商标图像检索方法研究
6.A moving target recognition and tracking method based on NMI feature
一种应用于大型综合实验的运动目标识别与跟踪方法
7.Image Retrieval Method Based on NMI Invariable Feature
基于NMI不变特征的图像检索方法
8.Target Recognition and Tracking Based on NMI Feature
基于NMI特征的目标识别与跟踪