Open D1vt opened 5 years ago
nice! now I will like to see some plots about the average error of the pose estimation using ideal points vs using random points.
the results using the code with adam
Cool, is this using the condition number of the big matrix or the condition number of the covariance matrix?
The condition number of the whole matrix. I am going to make one also for the covariance matrix.
On Thursday, April 4, 2019, Raul Acuna notifications@github.com wrote:
Cool, is this using the condition number of the big matrix or the condition number of the covariance matrix?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/D1vt/DLT3D/issues/13#issuecomment-479909875, or mute the thread https://github.com/notifications/unsubscribe-auth/Ar_mWEFvDk-1mL5TAFfbe7f58VjqS6qFks5vdgXngaJpZM4cRXHB .
-- http://uniqueminds.gr/ Diana Voutyrakou Founder, General Manager +306983456390|LinkedIn https://gr.linkedin.com/in/dialekti-athina-voutyrakou-646331107
Condition number of the cov matrix:
The code also proved that the points config. seems like a triangular prism:
Points are moving as close as they are able to the perimeter of the sphere: