【视觉slam14讲】高斯牛顿法

代码注释

#include <iostream>
#include <chrono>
#include <opencv2/opencv.hpp>
#include <Eigen/Core>
#include <Eigen/Dense>

using namespace std;
using namespace Eigen;

int main(int argc, char **argv) {
  double ar = 1.0, br = 2.0, cr = 1.0;         // 真实参数值，后面用来模拟观测数据
  double ae = 2.0, be = -1.0, ce = 5.0;        // 估计参数值，模拟slam中的估计值
  int N = 100;                                 // 数据点
  double w_sigma = 1.0;                        // 噪声Sigma值
  double inv_sigma = 1.0 / w_sigma;
  cv::RNG rng;                                 // OpenCV随机数产生器


std::cout<<"suijishu"<<rng()<<std::endl;

  vector<double> x_data, y_data;      // 数据,x 是自变量，y 是观测值
  for (int i = 0; i < N; i++) {
    double x = i / 100.0;
    x_data.push_back(x);
    //rng.gaussian(sigma) 会生成一个均值为 0、标准差为 sigma 的高斯分布随机数。在这里，标准差 w_sigma 被用于控制噪声的强度。
    //对应y值是一个真值+噪声值
    y_data.push_back(exp(ar * x * x + br * x + cr) + rng.gaussian(w_sigma * w_sigma));
    std::cout<<"随机数据"<<"   x_data   "<<x_data[i]<<"   y_data   "<<y_data[i]<<std::endl;
  }

  // 开始Gauss-Newton迭代
  int iterations = 100;    // 迭代次数
  double cost = 0, lastCost = 0;  // 本次迭代的cost和上一次迭代的cost

  chrono::steady_clock::time_point t1 = chrono::steady_clock::now();//记录时间
  for (int iter = 0; iter < iterations; iter++) {

//3*3的矩阵，初步设置所有元素为0
    Matrix3d H = Matrix3d::Zero();             // Hessian = J^T W^{-1} J in Gauss-Newton
    Vector3d b = Vector3d::Zero();             // bias
    cost = 0;

    for (int i = 0; i < N; i++) {
      double xi = x_data[i], yi = y_data[i];  // 第i个数据点，y是带噪声的
      double error = yi - exp(ae * xi * xi + be * xi + ce);//误差
      Vector3d J; // 雅可比矩阵
      //下面是求导
      J[0] = -xi * xi * exp(ae * xi * xi + be * xi + ce);  // de/da
      J[1] = -xi * exp(ae * xi * xi + be * xi + ce);  // de/db
      J[2] = -exp(ae * xi * xi + be * xi + ce);  // de/dc


//+=是累计赋值的意思
      H += inv_sigma * inv_sigma * J * J.transpose();
      b += -inv_sigma * inv_sigma * error * J;

      cost += error * error;
    }

    // 求解线性方程 Hx=b（高斯牛顿法的增量方程）
    Vector3d dx = H.ldlt().solve(b);
    if (isnan(dx[0])) {
      cout << "result is nan!" << endl;
      break;
    }

//表示当前代价函数值大于上一次的代价函数值输出警告信息并中断函数（可能陷入了局部最小值）
    if (iter > 0 && cost >= lastCost) {
      cout << "cost: " << cost << ">= last cost: " << lastCost << ", break." << endl;
      break;
    }

    ae += dx[0];
    be += dx[1];
    ce += dx[2];

    lastCost = cost;

    cout << "total cost: " << cost << ", \t\tupdate: " << dx.transpose() <<
         "\t\testimated params: " << ae << "," << be << "," << ce << endl;
  }

  chrono::steady_clock::time_point t2 = chrono::steady_clock::now();
  chrono::duration<double> time_used = chrono::duration_cast<chrono::duration<double>>(t2 - t1);
  cout << "solve time cost = " << time_used.count() << " seconds. " << endl;

  cout << "estimated abc = " << ae << ", " << be << ", " << ce << endl;
  return 0;
}

cmakelists

cmake_minimum_required(VERSION 2.8)
project(ch6)

set(CMAKE_BUILD_TYPE Release)
set(CMAKE_CXX_FLAGS "-std=c++14 -O3")

list(APPEND CMAKE_MODULE_PATH ${PROJECT_SOURCE_DIR}/cmake)

# OpenCV
find_package(OpenCV REQUIRED)
include_directories(${OpenCV_INCLUDE_DIRS})

# Ceres
find_package(Ceres REQUIRED)
include_directories(${CERES_INCLUDE_DIRS})



# Eigen
include_directories("/usr/include/eigen3")

add_executable(gaussNewton gaussNewton.cpp)
target_link_libraries(gaussNewton ${OpenCV_LIBS})