深度图像分割中的面元优化描述  

OPTIMAL DESCRIPTION OF SURFACE PRIMITIVE FOR RANGE IMAGE SEGMENTATION

作　　者：王国宇[1] 姬光荣[1] 李欣[1] 郑冰[1] 张海燕[1] 

机构地区：[1]青岛海洋大学电子工程系,青岛266003

出　　处：《电子与信息学报》2002年第11期1454-1461,共8页Journal of Electronics & Information Technology

摘　　要：该文提出了一种新的参数化面元的优化描述方法,基于二次型面表示模型,物体表面参数由深度图像数据拟合得到,基于一种改进的纠偏曲面拟合方法,提出了一个正态分布概率模型来描述面元估计的统计性质,并显式定义了面元描述的可靠性度量,以面元提取为目的,提出了一种新的基于参数估计的深度图像分割方法,在建立区域一致性判别优化准则的同时,使得区域增长总是收敛于所提取面元的“最可靠”描述,文中给出了实际深度图像分割的实验结果。This paper proposes a new algorithm of optimal description of surface primitive. The 3-D surface parameters are estimated through fitting the range data to a quadric model. Based on an improved bias-corrected fitting algorithm, a probabilistic model with normal pdf to describe the statistical characteristics of the surface estimate is proposed. Consequently a metric defined as 'measure of reliability' of the surface estimate is given. Towards reliable primitive extraction, a new method of estimation-based range image segmentation is thereby described. While the region homogeneity is defined on an optimal criterion, the region growing always converges to the state of 'the most reliable description' of the extracted surface primitives. The experimental results with real range images are given.

关 键 词：深度图像分割 面元 优化描述 参数估计 不确定性 区域增长 

分 类 号：TN911.73[电子电信—通信与信息系统]