search

TypingCat / spatial-topology-teleoperation

Teleoperation for Mobile Robot using Semantics
GNU General Public License v3.0
4 stars 0 forks source link

Clustering intersections #17

Closed TypingCat closed 3 years ago

TypingCat commented 3 years ago

Test #13 showed the possibility of the clustering approach. However, the current structure trains a new clustering model every time for the last 30 samples. Even there are no intersection angle. It need to be improved.

TypingCat commented 3 years ago

Principal Component Analysis(PCA) results are visualized by ellipse in commit 01d010da3f7fad831cb64d56fd24966ec69150e2.

All clusters store last 300 samples. For each cluster, mean and standard deviation are calculated by PCA. It represented by center and size of an ellipse. In the following, intersections are represented by 6 ellipses. Distance errors from the robot tend to be larger than angle errors.

intersection_cluster_pca

TypingCat commented 3 years ago

Pause this issue for modularization. See #18.

TypingCat commented 3 years ago

The clustering module was tested on real robot. This robot moves at 0.2m/s, and its path is exactly same as https://github.com/TypingCat/waffle/issues/17#issuecomment-788610727. Intersection cluster covariance were represented by yellow cylinder. All intersections were detected. The problem is the misdetection. The size of covariance or the number of samples can be a clue.

intersection_clustering_ros