lm opencv 算法_Levenberg–Marquardt算法学习（和matlab的LM算法对比）
回顾高斯牛顿算法，引入LM算法
惩罚因子的计算(迭代步子的计算)
完整的算法流程及代码样例
1. 回顾高斯牛顿，引入LM算法 根据之前的博文：Gauss-Newton算法学习
假设我们研究如下形式的非线性最小二乘问题：
r(x)为某个问题的残差residual，是关于x的非线性函数。我们知道高斯牛顿法的迭代公式：
Levenberg–Marquardt算法是对高斯牛顿的改进，在迭代步长上略有不同：
最速下降法对初始点没有特别要求具有整体收敛性，但是相邻两次的搜索方向是相互垂直的，所以收敛并不一定快。总而言之就是：当目标函数的等值线接近于圆(球)时，下降较快；等值线类似于扁长的椭球时，一开始快，后来很慢。This is good if the current iterate is far from the solution.
c. 如果μ的值很小，那么hlm成了高斯牛顿法的方向(适合迭代的最后阶段，非常接近最优解，避免了最速下降的震荡)
由此可见，惩罚因子既下降的方向又影响下降步子的大小。
2. 惩罚因子的计算[迭代步长计算]我们的目标是求f的最小值，我们希望迭代开始时，惩罚因子μ被设定为较小的值，若
信赖域方法与线搜索技术一样,也是优化算法中的一种保证全局收敛的重要技术. 它们的功能都是在优化算法中求出每次迭代的位移, 从而确定新的迭代点.所不同的是: 线搜索技术是先产生位移方向(亦称为搜索方向), 然后确定位移的长度(亦称为搜索步长)。而信赖域技术则是直接确定位移, 产生新的迭代点。
信赖域方法的基本思想是:首先给定一个所谓的“信赖域半径”作为位移长度的上界，并以当前迭代点为中心以此“上界”为半径确定一个称之为“信赖域”的闭球区域。然后,通过求解这个区域内的“信赖域子问题”(目标函数的二次近似模型) 的最优点来确定“候选位移”。若候选位移能使目标函数值有充分的下降量, 则接受该候选位移作为新的位移,并保持或扩大信赖域半径, 继续新的迭代。否则, 说明二次模型与目标函数的近似度不够理想,需要缩小信赖域半径，再通过求解新的信赖域内的子问题得到新的候选位移。如此重复下去，直到满足迭代终止条件。 现在用信赖域方法解决之前的无约束线性规划：
如果q很大，说明L(h)非常接近F(x+h)，我们可以减少惩罚因子μ，以便于下次迭代此时算法更接近高斯牛顿算法。如果q很小或者是负的，说明是poor approximation，我们需要增大惩罚因子，减少步长，此时算法更接近最速下降法。具体来说，
a.当q大于0时，此次迭代有效：
b．当q小于等于0时，此次迭代无效：
3.完整的算法流程及代码距离
LM的算法流程和高斯牛顿几乎一样，只是迭代步长求法利用信赖域法
(1)给定初始点x(0)，允许误差ε>0，置k=0
(2)当f(xk+1)-f(xk)小于阈值ε时，算法退出，否则(3)
(3)xk+1=xk+hlm，代入f，返回(1)
两个例子还是沿用之前的。
例子1，根据美国1815年至1885年数据，估计人口模型中的参数A和B。如下表所示，已知年份和人口总量，及人口模型方程，求方程中的参数。
// A simple demo of Gauss-Newton algorithm on a user defined function
#include
#include
#include
usingnamespacestd;
usingnamespacecv;
constdoubleDERIV_STEP = 1e-5;
constintMAX_ITER = 100;
voidLM(double(*Func)(constMat &input,constMat ¶ms),// function pointer
constMat &inputs,constMat &outputs, Mat ¶ms);
doubleDeriv(double(*Func)(constMat &input,constMat ¶ms),// function pointer
constMat &input,constMat ¶ms,intn);
// The user defines their function here
doubleFunc(constMat &input,constMat ¶ms);
intmain()
{
// For this demo we're going to try and fit to the function
// F = A*exp(t*B)
// There are 2 parameters: A B
intnum_params = 2;
// Generate random data using these parameters
inttotal_data = 8;
Mat inputs(total_data, 1, CV_64F);
Mat outputs(total_data, 1, CV_64F);
//load observation data
for(inti=0; i
inputs.at(i,0) = i+1;//load year
}
//load America population
outputs.at(0,0)= 8.3;
outputs.at(1,0)= 11.0;
outputs.at(2,0)= 14.7;
outputs.at(3,0)= 19.7;
outputs.at(4,0)= 26.7;
outputs.at(5,0)= 35.2;
outputs.at(6,0)= 44.4;
outputs.at(7,0)= 55.9;
// Guess the parameters, it should be close to the true value, else it can fail for very sensitive functions!
Mat params(num_params, 1, CV_64F);
//init guess
params.at(0,0) = 6;
params.at(1,0) = 0.3;
LM(Func, inputs, outputs, params);
printf("Parameters from GaussNewton: %lf %lf\n", params.at(0,0), params.at(1,0));
return0;
}
doubleFunc(constMat &input,constMat ¶ms)
{
// Assumes input is a single row matrix
// Assumes params is a column matrix
doubleA = params.at(0,0);
doubleB = params.at(1,0);
doublex = input.at(0,0);
returnA*exp(x*B);
}
//calc the n-th params' partial derivation ， the params are our final target
doubleDeriv(double(*Func)(constMat &input,constMat ¶ms),constMat &input,constMat ¶ms,intn)
{
// Assumes input is a single row matrix
// Returns the derivative of the nth parameter
Mat params1 = params.clone();
Mat params2 = params.clone();
// Use central difference to get derivative
params1.at(n,0) -= DERIV_STEP;
params2.at(n,0) += DERIV_STEP;
doublep1 = Func(input, params1);
doublep2 = Func(input, params2);
doubled = (p2 - p1) / (2*DERIV_STEP);
returnd;
}
voidLM(double(*Func)(constMat &input,constMat ¶ms),
constMat &inputs,constMat &outputs, Mat ¶ms)
{
intm = inputs.rows;
intn = inputs.cols;
intnum_params = params.rows;
Mat r(m, 1, CV_64F); // residual matrix
Mat r_tmp(m, 1, CV_64F);
Mat Jf(m, num_params, CV_64F); // Jacobian of Func()
Mat input(1, n, CV_64F); // single row input
Mat params_tmp = params.clone();
doublelast_mse = 0;
floatu = 1, v = 2;
Mat I = Mat::ones(num_params, num_params, CV_64F);//construct identity matrix
inti =0;
for(i=0; i
doublemse = 0;
doublemse_temp = 0;
for(intj=0; j
for(intk=0; k
input.at(0,k) = inputs.at(j,k);
}
r.at(j,0) = outputs.at(j,0) - Func(input, params);//diff between previous estimate and observation population
mse += r.at(j,0)*r.at(j,0);
for(intk=0; k
Jf.at(j,k) = Deriv(Func, input, params, k);//construct jacobian matrix
}
}
mse /= m;
params_tmp = params.clone();
Mat hlm = (Jf.t()*Jf + u*I).inv()*Jf.t()*r;
params_tmp += hlm;
for(intj=0; j
r_tmp.at(j,0) = outputs.at(j,0) - Func(input, params_tmp);//diff between current estimate and observation population
mse_temp += r_tmp.at(j,0)*r_tmp.at(j,0);
}
mse_temp /= m;
Mat q(1,1,CV_64F);
q = (mse - mse_temp)/(0.5*hlm.t()*(u*hlm-Jf.t()*r));
doubleq_value = q.at(0,0);
if(q_value>0)
{
doubles = 1.0/3.0;
v = 2;
mse = mse_temp;
params = params_tmp;
doubletemp = 1 - pow(2*q_value-1,3);
if(s>temp)
{
u = u * s;
}else
{
u = u * temp;
}
}else
{
u = u*v;
v = 2*v;
params = params_tmp;
}
// The difference in mse is very small, so quit
if(fabs(mse - last_mse)
break;
}
//printf("%d: mse=%f\n", i, mse);
printf("%d %lf\n", i, mse);
last_mse = mse;
}
}
A=7.0,B=0.26(初始值，A=6,B=0.3)，100次迭代到第7次收敛了。和之前差不多，但是LM对于初始的选择是非常敏感的，如果A=6，B=6，则拟合失败！
我调用了matlab的接口跑LM，结果也是一样错误的，图片上可以看到拟合失败
clc;
clear;
a0=[6,6];
options=optimset('Algorithm','Levenberg-Marquardt','Display','iter');
data_1=[1 2 3 4 5 6 7 8];
obs_1=[8.3 11.0 14.7 19.7 26.7 35.2 44.4 55.9];
a=lsqnonlin(@myfun,a0,[],[],options,data_1,obs_1);
plot(data_1,obs_1,'o');
hold on
plot(data_1,a(1)*exp(a(2)*data_1),'b');
plot(data_1,7*exp(a(2)*data_1),'b');
%hold off
a(1)
a(2)
function E = myfun(a, x,y)
%这是一个测试文件用于测试 lsqnonlin
% Detailed explanation goes here
x=x(:);
y=y(:);
Y=a(1)*exp(a(2)*x);
E=y-Y;
end
最后一个点拟合失败的，所以函数不对的
因此虽然莱文博格-马夸特迭代法能够自适应的在高斯牛顿和最速下降法之间调整，既可保证在收敛较慢时迭代过程总是下降的，又可保证迭代过程在解的邻域内迅速收敛。但是，LM对于初始点选择还是比较敏感的！
例子2：我想要拟合如下模型，
由于缺乏观测量，就自导自演，假设4个参数已知A=5,B=1,C=10,D=2，构造100个随机数作为x的观测值，计算相应的函数观测值。然后，利用这些观测值，反推4个参数。
// A simple demo of Gauss-Newton algorithm on a user defined function
#include
#include
#include
usingnamespacestd;
usingnamespacecv;
constdoubleDERIV_STEP = 1e-5;
constintMAX_ITER = 100;
voidLM(double(*Func)(constMat &input,constMat ¶ms),// function pointer
constMat &inputs,constMat &outputs, Mat ¶ms);
doubleDeriv(double(*Func)(constMat &input,constMat ¶ms),// function pointer
constMat &input,constMat ¶ms,intn);
// The user defines their function here
doubleFunc(constMat &input,constMat ¶ms);
intmain()
{
// For this demo we're going to try and fit to the function
// F = A*sin(Bx) + C*cos(Dx)
// There are 4 parameters: A, B, C, D
intnum_params = 4;
// Generate random data using these parameters
inttotal_data = 100;
doubleA = 5;
doubleB = 1;
doubleC = 10;
doubleD = 2;
Mat inputs(total_data, 1, CV_64F);
Mat outputs(total_data, 1, CV_64F);
for(inti=0; i
doublex = -10.0 + 20.0* rand() / (1.0 + RAND_MAX);// random between [-10 and 10]
doubley = A*sin(B*x) + C*cos(D*x);
// Add some noise
// y += -1.0 + 2.0*rand() / (1.0 + RAND_MAX);
inputs.at(i,0) = x;
outputs.at(i,0) = y;
}
// Guess the parameters, it should be close to the true value, else it can fail for very sensitive functions!
Mat params(num_params, 1, CV_64F);
params.at(0,0) = 1;
params.at(1,0) = 1;
params.at(2,0) = 8;// changing to 1 will cause it not to find the solution, too far away
params.at(3,0) = 1;
LM(Func, inputs, outputs, params);
printf("True parameters: %f %f %f %f\n", A, B, C, D);
printf("Parameters from GaussNewton: %f %f %f %f\n", params.at(0,0), params.at(1,0),
params.at(2,0), params.at(3,0));
return0;
}
doubleFunc(constMat &input,constMat ¶ms)
{
// Assumes input is a single row matrix
// Assumes params is a column matrix
doubleA = params.at(0,0);
doubleB = params.at(1,0);
doubleC = params.at(2,0);
doubleD = params.at(3,0);
doublex = input.at(0,0);
returnA*sin(B*x) + C*cos(D*x);
}
//calc the n-th params' partial derivation ， the params are our final target
doubleDeriv(double(*Func)(constMat &input,constMat ¶ms),constMat &input,constMat ¶ms,intn)
{
// Assumes input is a single row matrix
// Returns the derivative of the nth parameter
Mat params1 = params.clone();
Mat params2 = params.clone();
// Use central difference to get derivative
params1.at(n,0) -= DERIV_STEP;
params2.at(n,0) += DERIV_STEP;
doublep1 = Func(input, params1);
doublep2 = Func(input, params2);
doubled = (p2 - p1) / (2*DERIV_STEP);
returnd;
}
voidLM(double(*Func)(constMat &input,constMat ¶ms),
constMat &inputs,constMat &outputs, Mat ¶ms)
{
intm = inputs.rows;
intn = inputs.cols;
intnum_params = params.rows;
Mat r(m, 1, CV_64F); // residual matrix
Mat r_tmp(m, 1, CV_64F);
Mat Jf(m, num_params, CV_64F); // Jacobian of Func()
Mat input(1, n, CV_64F); // single row input
Mat params_tmp = params.clone();
doublelast_mse = 0;
floatu = 1, v = 2;
Mat I = Mat::ones(num_params, num_params, CV_64F);//construct identity matrix
inti =0;
for(i=0; i
doublemse = 0;
doublemse_temp = 0;
for(intj=0; j
for(intk=0; k
input.at(0,k) = inputs.at(j,k);
}
r.at(j,0) = outputs.at(j,0) - Func(input, params);//diff between estimate and observation population
mse += r.at(j,0)*r.at(j,0);
for(intk=0; k
Jf.at(j,k) = Deriv(Func, input, params, k);//construct jacobian matrix
}
}
mse /= m;
params_tmp = params.clone();
Mat hlm = (Jf.t()*Jf + u*I).inv()*Jf.t()*r;
params_tmp += hlm;
for(intj=0; j
r_tmp.at(j,0) = outputs.at(j,0) - Func(input, params_tmp);//diff between estimate and observation population
mse_temp += r_tmp.at(j,0)*r_tmp.at(j,0);
}
mse_temp /= m;
Mat q(1,1,CV_64F);
q = (mse - mse_temp)/(0.5*hlm.t()*(u*hlm-Jf.t()*r));
doubleq_value = q.at(0,0);
if(q_value>0)
{
doubles = 1.0/3.0;
v = 2;
mse = mse_temp;
params = params_tmp;
doubletemp = 1 - pow(2*q_value-1,3);
if(s>temp)
{
u = u * s;
}else
{
u = u * temp;
}
}else
{
u = u*v;
v = 2*v;
params = params_tmp;
}
// The difference in mse is very small, so quit
if(fabs(mse - last_mse)
break;
}
//printf("%d: mse=%f\n", i, mse);
printf("%d %lf\n", i, mse);
last_mse = mse;
}
}
我们看到迭代了100次，结果几何和高斯牛顿算出来是一样的。我们绘制LM和高斯牛顿的残差函数收敛过程，发现LM一直是总体下降的，没有太多反复。
高斯牛顿：
LM：