Local Feature Extraction Using Scale-Space Decomposition

In our recent work we have introduced a framework for extracting features from solid of mechanical artifacts in polyhedral representation based on scale-space feature decomposition [1]. Our approach used recent developments in efficient hierarchical decomposition of metric data using its spectral properties. In that work, through spectral decomposition, we were able to reduce the problem of matching to that of computing a mapping and distance measure between vertex-labeled rooted trees. This work discusses how Scale-Space decomposition frame-work could be extended to extract features from CAD models in polyhedral representation in terms of surface triangulation. First, we give an overview of the Scale-Space decomposition approach that is used to extract these features. Second, we discuss the performance of the technique used to extract features from CAD data in polyhedral representation. Third, we show the feature extraction process on noisy data — CAD models that were constructed using a 3D scanner. Finally, we conclude with discussion of future work.