【Hector slam】A Flexible and Scalable SLAM System with Full 3D Motion Estimation

做者總結了SLAM前端和後端的區別

While SLAM frontends are used to estimate robot movement online in real-time,

the backend is used to perform optimization of the pose graph given constraints

between poses that have been generated before using the frontend. 

前端 用來在線實時估計機器人運動，

後端用來優化位姿

而這篇文章服務於SLAM前端，並不提供後端優化。能夠估計6DOF呢。

our system has to estimate the full 6DOF state consisting  of translation and rotation of the platform.

To achieve this,the system consists of two major components.

A navigation filter fuses information from the inertial measurement unit

and other available sensors to form a consistent 3D solution,

while a 2D SLAM system is used to provide position and

heading information within the ground plane.

********　　　我關注的重點在2D slam

數據點的預處理是必不可少的

A 而後建圖：雙線性濾波估計佔據柵格的機率

 

B 幀間匹配：高斯牛頓方法，不須要創建點之間的關係

Our approach is based on optimization of the alignment of beam endpoints with the map learnt so far.

The basic idea using a Gauss-Newton approach is inspired by work in computer vision

[19 An iterative image registration technique with an application to stereo vision (darpa)].

Using this approach, there is no need for a data association search between beam endpoints or an exhaustive pose search.

As scans get aligned with the existing map, the matching is implicitly performed with all preceding scans

 

C 多分辨率地圖表示

像圖像金字塔同樣，幀匹配的時候，從最粗的地圖開始，結果做爲下一精度匹配的初始估計。

The scan alignment process is started at the coarsest map level, with the resulting
estimated pose getting used as the start estimate for the next level