Unique signatures of histograms for local surface description (original) (raw)

The research explores unique signature representations of histograms for effectively describing local surface geometries in 3D space. It highlights the evolution of 3D surface matching techniques, emphasizing the shift from global to local approaches involving geometric descriptors. The proposed taxonomy categorizes existing descriptor methods into signature and histogram-based approaches, illustrating their respective strengths and weaknesses in tasks such as object recognition and surface alignment.