i am working on modelling to robot environment.
I am using Sick Laser Data to get a point cloud for
robot environment modelling
i want to extract the
geomtrical features from the point cloud . I am
converting the point cloud in to a Range Image
and then applying some of my own image processing algorithms
for extraction of the features.
further after extracting the feature from point cloud I want to use is for SLAM
through a sequence of point clouds.
I just want to get if this has some value addition
in exiting techniques of feature extraction and SLAM from point cloud as this technique
is very fast .