3D volume reconstruction [closed]
Hello,
Well i have a little problem and need some tips to solve it.
So, i have a stereo camera point cloud data, and i need to cluster it. It is very simple cluster the data directly from 3d using euclidean cluster extraction. However that is not my goal. I have also a laser scan pointed on the same direction with the camera and the laser gives me more believable information about clusters, however it is only 2D. The idea here is clustering the 2D laser scan with euclidean cluster extraction and then find a way to seed these points on stereo camera point cloud and then search for the neighbor points to estimate the size of this clusters. I looked on octree and kdtree search methods, but it is not what i needed. the octree gives me all the points inside a given radius from a seed point, and what i want is to see how long the cluster goes with a given threshold. Confusing?
anybody know a way to do that?
You seem to want euclidean clustering, and say so, but then also seem to say that euclidean clustering is not what you want. Could you please clarify?