Graph Cuts算法实现过程中的关键问题研究  

The Study of Key Issues for Graph Cuts Realization

作　　者：张晓峰[1] 

机构地区：[1]南通大学计算机科学与技术学院,江苏南通226019

出　　处：《软件导刊》2014年第12期44-45,共2页Software Guide

摘　　要：Graph Cuts(GC)算法是一个基于图论的交互式目标提取算法,该方法将图像建模为一个区域和边缘的约束模型,通过求解该模型的最小割获得一个优化的目标分割边界。在使用程序实现该算法时会遇到种种问题,详细叙述了GC算法的实现过程,为进一步研究GC相关方法奠定基础。Graph Cuts （GC） algorithm is an interactive method for object detection based on Graph theory .This algorithm translates image segmentation into a model which is composed of region properties and boundary properties .Then a min cut is obtained by a optimal algorithm for object segmentation .Because the deep theory of GC algorithm ,many difficulties will occur in its realization .In this paper ,the realization of GC algorithm is clearly described ,which make a fundament for further research of GC model .

关 键 词：目标提取 GRAPH CUTS 优化分割 

分 类 号：TP303[自动化与计算机技术—计算机系统结构]