最新最小二乘非线性优化库，实测可用，兼容vs2013 ，Maximum likelihood estimation (MLE) is a well-known estimation method used in many robotic and computer vision applications. Under Gaussian assumption, the MLE converts to a nonlinear least squares (NLS) problem. Efficient solutions to NLS exist and they are based on iteratively solving sparse linear systems until convergence.

姿势图优化 使用Ceres Solver进行2D姿态图优化的示例代码。 依存关系 本征3.3或更高版本 Ceres Solver 1.12.0或更高版本 Gflags 2.2.0或更高版本 带有matplotlib的Python 建造 $ git clone https://github.com/shinsumicco/pose-graph-optimization.git $ cd pose-graph-optimization $ mkdir build $ cd build $ cmake .. $ make -j4 优化 $ cd pose-graph-optimization/build $ bin/se2_optimize --filename ../sample/manhattan.g2o Number of poses: 3500 Number of constraints

ceres solver库+ceres solver的使用

https://ceres-solver.googlesource.com/ceres-solver.git ， version 1.13.0

ceres库已通过编译和配置（VS2019+win10）debug版（不包括release版）

麻省理工大学开源数据集，格式为.g2o文件，包含2D Pose Graph Optimization与3D Pose Graph Optimization数据，可以使用ceres 示例程序pose_graph运行

window10+vs2019 环境下编译ceres库，文件包含所有依赖的库的源码，以及编译后的Debug库、Release库

windows10 + vs 2019环境下编译后的ceres库，包含lib、bin以及头文件

包中包含ceres库的几个源码，包括ceres solver,eigen,gflags, glogs等库，和在win10+vs2017平台下编译好了的include和lib文件，包括Release版本和Debug版本。已经测试都可用。

Sophus + ceres 对SE3的位姿优化实践，PnP iter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time 0 6.971092e+01 0.00e+00 2.29e+01 0.00e+00 0.00e+00 1.00e+04 0 7.45e-05 2.28e-04 1 5.852248e+01 1.12e+01 8.91e+00 1.03e+01 1.82e+00 3.00e+04 0 1.41e-04 4.02e-04 2 2.764596e+01 3.09e+01 2.00e+01 2.52e+00 4.94e+00 9.00e+04 0 7.64e-05 4.94e-04 3 7.834946e-02 2.76e+01 1.37e+01 2.36e-01 4.48e+00 2.70e+05 0 7.05e-05 5.76e-04 4 4.140774e-02 3.69e-02 1.03e+01 2.52e-03 1.09e+00 8.10e+05 0 6.93e-05 6.57e-04 5 4.119084e-02 2.17e-04 1.67e+01 2.11e-04 1.07e+00 2.43e+06 0 6.94e-05 7.40e-04 6 4.118984e-02 1.00e-06 1.63e+00 1.41e-05 1.06e+00 7.29e+06 0 6.83e-05 8.18e-04 7 4.118983e-02 4.19e-09 1.06e-01 9.14e-07 1.06e+00 2.19e+07 0 6.74e-05 8.94e-04 Ceres Solver Report: Iterations: 8, Initial cost: 6.971092e+01, Final cost: 4.118983e-02, Termination: CONVERGENCE after optimize: 0.925421 -0.287927 0.246363 -0.722 0.336739 0.92301 -0.18617 -8.15989 -0.173792 0.255245 0.951129 9.10729 0 0 0 1 err_opt:0.0038567 -0.00193814 0.00258262 -0.0020643 -0.000398187 -0.000119575 0.000116657 true value: 0.925417 -0.288133 0.246137 -0.725046 0.336824 0.923045 -0.185843 -8.15777 -0.173648 0.254887 0.951251 9.10632 0 0 0 1 err_init:10.3912 -3.06362 8.02512 -5.81668 -0.14186 0.155837 -0.557138