/*
这是一篇SLAM的“hello world”说明书,结构框架介绍得很全面,可以建立初步的认识。关于SLAM,还有后续,更具体更常见的例子,慢慢看,慢慢学。
*/
SLAM是Simultaneous Localization and Mapping的缩写,即”即时定位与地图构建 “。
It helps robots construct their surrounding map and guide themselves.
SLAM needs a huge amount of hardware, and it’s more like a concept than an algorithm. EKF is a complete solution for SLAM( Extended Kalman Filter).
SLAM consists of multiple parts: Landmark extraction, data association, state
estimation, state update and landmark update.( 地标提取,数据关联,状态估计,状态更新和地标更新)
To do SLAM there is the need for hardware like
1.a mobile robot (with wheels and odometry performance)
2. a range measurement device.( laser scanner + sonar + vision)
Steps:
A slam process
a) Landmarks should be easily re-observable.
b) Individual landmarks should be distinguishable from each other.
c) Landmarks should be plentiful in the environment.
d) Landmarks should be stationary.
The implementation of the EKF and other codes.
只言片语 随手摘录
以上。