基于控制点的二值图像边界矢量化算法  

Control Point Based Algorithm of Boundary Vectorization in a Binary Image

作　　者：张显全[1] 唐振军[1] 王继军[1] 

机构地区：[1]广西师范大学计算机科学系,桂林541004

出　　处：《计算机科学》2009年第1期169-170,189,共3页Computer Science

基　　金：广西自然科学基金(0447035);广西教育厅项目(200607MS135);广西研究生教育创新计划项目(200610602812M37)资助

摘　　要：提出一种二值图像边界矢量化算法,实现从图像到图形的自动转换。先通过边界跟踪算法获取二值图像的边界,并利用SUSAN方法计算出边界上的角点,根据距离关系计算角点间的普通控制点,最后运用角点和这些普通控制点的3次B样条曲线拟合边界。实验结果表明算法有效实现了图像矢量化,解决了边界矢量化问题,具有较强的实用性。We proposed an algorithm of boundary vectorization for binary images, which can convert an image into a graphic. The boundary tracing algorithm was exploited to extract the boundary of a binary image, and the SUSAN approach was also applied to detecting the corners on the boundary. In order to approximate the image boundary, the control points between two adjacent corners were determined by their offset distances. The cubic B-Spline curve was used to fit the boundary. The experimental results show that the proposed algorithm can convert a binary image into a graphic and solve theproblem of boundary vectorization. It is practical.

关 键 词：边界 矢量化 控制点 角点 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]