Dept. of Electr. and Comput. Eng., Michigan Univ., Dearborn, MI.
IEEE Trans Image Process. 1997;6(6):826-30. doi: 10.1109/83.585233.
A novel method for range image segmentation is presented in this paper. It is based on an integration of edge and region information. The algorithm consists of three steps: edge and critical point detection, triangulation, and region growing. Experimental results show that the method is efficient for segmentation of the range images that contain polyhedral objects. A three-dimensional (3-D) surface structure graph (SSG) obtained from the segmentation is a description of the surface structure about an object. Therefore, a segmentation result also presents a data set that can be used to establish a surface model for computer-aided-design-based (CAD-based) vision and object recognition.
本文提出了一种新的深度图像分割方法。它基于边缘和区域信息的集成。该算法包括三个步骤：边缘和关键点检测、三角剖分和区域生长。实验结果表明，该方法对于包含多面体物体的深度图像的分割是有效的。从分割中得到的三维（3-D）表面结构图（SSG）是关于物体表面结构的描述。因此，分割结果也提供了一个数据集，可用于建立基于计算机辅助设计（CAD）的视觉和目标识别的表面模型。
