乍一看代码Linear phase-height model和Polynomial phase-height model的step1几乎一样，但是略有不同
两者都是说的相位-高度模型下的，看论文可以更清晰地明了
Calibration of fringe projection profilometry: A comparative review
相同点：输入相同
不同点：在于n_ord
%Polynomial phase-height model
%% step 1: load the unwrapped phase of the moving plane at different heights
tic;
load_data = load('up_all.mat');
up_all = load_data.up_all;
ref_heights = load_data.ref_heights;
n_ord = 3; % polynomial fitting
img_width = 640; % image width
img_height = 480; % image height
%Linear phase-height model
%% step 1: load the unwrapped phase of the moving plane at different heights
tic;
load_data = load('up_all.mat');
up_all = load_data.up_all;
ref_heights = load_data.ref_heights;
n_ord = 1; % linear fitting
img_width = 640; % image width
img_height = 480; % image height
%Linear inverse phase-height model
%% step 1: load the unwrapped phase of the moving plane at different heights
tic;
load_data = load('up_all.mat');
up_all = load_data.up_all;
ref_heights = load_data.ref_heights;
img_width = 640; % image width
img_height = 480; % image height
而论文中的多项式相位-高度模型和基于多项式拟合的结构光系统标定完全不一样
论文提到以下五种模型，比较有价值的也就三角法了，学习中
线性相位高度模型
逆线性相位高度模型
多项式相位高度模型
基于控制方程的相位高度模型
三角立体模型
