用于深度图内3—D物体识别的表面特性

SURFACE CHARACTERIZATION FOR THREE-DIMENSIONAL OBJEC T RECOGNITION IN DEPTH MAPS

本文以精确的数学形式重新定义深度图物体的识别为广义道集合映射。由于怎样有效地计算此类映射还没有通用的理论,因此,我们提出基于物体表面特性和表面匹配的计算理论。此理论对深入研究与分析深度图物体识别是十分有用的。

This paper redefines depth map object recognition in precise mathematical terms as a generalized inverse set mapping. Since there is no general theory available on how to compute such a mapping efficiently,a computational theory based on surface characterization and surface matching is put forth. This theory is very useful to conduct in-depth depth research on;depth map object recognition.