Code for Slam project of SJTU Matlab Course.
[toc]
Feature matching codes are in Odometry/FeatureMatch
edit images path and run testORB.m to see the results of Feature detect of a giving image.
edit images path and run trial_2.m to see the results of Extract and Match Feature Points.
change the excepted number of Feature Points in para.m. change the Thresholds of Feature match in violent_match.m.
in matlab prompt
>> cd slamtb-graph/slamtb-ekf % or just double click to enter the folder
>> slamrc % load parameter
>> slamtb_graph % for slamtb-graph
% or
>> slamtb % for slamtb-ekf
- 视觉里程计
- 利用slamtb搭建地图
- 卡尔曼滤波
- 非线性优化
- 特征点匹配算法优化
the slamtb is downloaded from github: joansola
slamtb-graph & slamtb-ekf are two corresponding branches of this repo
To avoid unnecessary files in this repo, .mat files and all images are ignored, everything in figure folder are ignored, place images as you want.
With the hard work of hwk, qk, xjx, zzh in SJTU Matlab Course: (2019-2020-2)-ME391-1-Matlab及其工程应用