You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am trying to implement an algorithm ICP style. And to find correspondences between the two d-dimensional point sets (assuming same number of points per set).
However because I would like to enforce bijection in these correspondences my approach was to remove certain nodes from the tree iteratively.
Roughly the following pseudocode
input P,Q : d-dimensional point sets with |P| = |Q|
output: corresponce map C
T = KDTree constructed from Q
C = correspondence map of size |Q|
for each p in P:
j = T.find(p);
T.remove(j); //How to do this?
C[p.index()] = j;
This should guarantee the bijection. But by looking through past issues it doesn't seem to me you implemented such feature?
Can you tell me how to do this? or if removal is not available maybe some approaches maybe based on re-adjusting the tree cleverly?
I could construct a new tree everytime but I have a feeling at that point might be better just doing the naive way. But would be nice to have a solution that uses a KDTree.
The text was updated successfully, but these errors were encountered:
Hi,
I am trying to implement an algorithm ICP style. And to find correspondences between the two d-dimensional point sets (assuming same number of points per set).
However because I would like to enforce bijection in these correspondences my approach was to remove certain nodes from the tree iteratively.
Roughly the following pseudocode
This should guarantee the bijection. But by looking through past issues it doesn't seem to me you implemented such feature?
Can you tell me how to do this? or if removal is not available maybe some approaches maybe based on re-adjusting the tree cleverly?
I could construct a new tree everytime but I have a feeling at that point might be better just doing the naive way. But would be nice to have a solution that uses a KDTree.
The text was updated successfully, but these errors were encountered: