next up previous
Next: Competitive Object Learning Up: Object Localization Previous: Model Matching

Evaluating the Match

Generally, one is interested in the quality of the matching, i.e., the accuracy of the model pose inside the 3D data. Many application specific tasks require this estimation, e.g., complicated robot navigation tasks. However, the value of the error function ([*]) does not give this information, since point densities influence this value. Different point densities are the result of the scanning process, i.e., the spherical and continuous measurement of the laser. The scanner emits the laser beams in a spherical way such that the data points close to the source are more dense. A competitive learning technique is used to subsample the model and data set.


root 2005-05-03