Title :
A Novel Shape Descriptor: Gaussian Curvature Moment Invariants
Author :
Guo Kehua ; Li Min
Author_Institution :
Sch. of Inf. Sci. & Technol., Central South Univ., Changsha, China
Abstract :
The moment descriptor is combined with Gaussian curvature for three-dimensional shape representation and a novel three-dimensional shape descriptor combined local with global representations is proposed in this paper. Normalization process to the new moment invariants is presented and their independence to the translation, rotation and scaling transforms is proved. Experiments indicate a better classification result to objects with slight different shape characteristic compared with some traditional approaches without increasing the running complexity.
Keywords :
Gaussian processes; computational geometry; image classification; object recognition; shape recognition; Gaussian curvature moment invariants; classification result; global representations; moment descriptor; moment invariants; normalization process; rotation; running complexity; scaling transforms; shape descriptor; three-dimensional shape representation; translation transforms; Character recognition; Image segmentation; Information science; Kernel; MPEG 7 Standard; Moment methods; Noise shaping; Pattern recognition; Robustness; Shape measurement;
Conference_Titel :
Information Science and Engineering (ICISE), 2009 1st International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4244-4909-5
DOI :
10.1109/ICISE.2009.125