【摘要】 提出一种基于拓扑图及三维特征的三维物体识别方法.该方法在图象预处理阶段先检测物体可视表面的闭合边界,再依据闭合边界对物体进行表面分割,并用一个景物特征图对景物进行表达(SAGR).在建模阶段,先采用多视图表面模型方法建模,再依据表面分割,对模型用一个模型特征图进行表达(MAGR).在物体识别阶段,将景物与模型的匹配分3部分进行,即扫描部件、图的匹配部件及匹配检验部件.匹配首先是基于 SAGR 与 MAGR 图的拓扑性质的匹配,然后是基于三维特征的匹配.根据匹配结果,或对未知物体自动建模,或给出识别结果.更多还原

【Abstract】 A new 3-D objects recognition approach based on topology and 3-D attributes is presented.In the preprocessing stage,the enclosed edges of visible object surfaces are detected first and a scene attribute graph to represent the scene(SAGR)is used.In modeling stage,a multiple-view surface model is applied to build a model represented by a model attribute graph(MAGR)accord- ing to surface segmentation.In object recognition stage,the approach divides the matching between scene and model into three parts,i.e.the Scanner,the Graph Marcher and the Matching Verifier.The matching is based on the topology at first,then on the 3-D attributes between SAGR and MAGR.Aecording to the matching the approach either automatically builds model for unknown scene or gives the recognition results.The effectiveness of the approach presented in this paper has been shown in the experiments.更多还原