基于三维上下文的图像几何分割方法  

Image Geometric Segmentation Method Based on 3D Context

作　　者：陈肖[1] 王敏[1] 陶铭洋 

机构地区：[1]河海大学计算机与信息学院,江苏南京211100

出　　处：《计算机与现代化》2014年第4期38-40,46,共4页Computer and Modernization

摘　　要：图像分割是计算机视觉中最基本的前期准备工作,也是由图像处理到图像分析的关键步骤。传统图像分割方法得到的结果往往是不具有几何意义的独立区域,而现有的几何分割方法分割准确率较低。本文提出一种基于三维上下文的场景图像几何分割算法,将场景图像的三维上下文信息成功应用到图像的几何分割中。实验表明,利用这种方法对图像进行几何分割具有较高的准确率。Image segmentation is the most basic preparatory work in computer vision, and it is the key step from image processing to image analysis. The result of image segmentation is always separate areas without geometric meaning, and the accuracies of ex- isting methods of geometry segmentation are not very high. In this paper, an image geometric segmentation method is proposed and 3D context information is successfully applied to geometric segmentation of scene images. Experiments show that this method can do geometric segmentation with higher accuracy.

关 键 词：图像分割 几何分割 三维上下文 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]