摘要
提出一种应用于图像表达中的轮廓层次结构树。与Fub提出的区域包含关系树相比,本方法增加了对相邻轮廓所构成的整体轮廓的表示,对图像轮廓关系的描述更为准确完整,可以用在多层次图像重构和图像匹配等图像表达的应用中。实验表明,该方法在图像表达应用中有较好的效果。
A contour-based hierarchical tree structure used in image representation was discussed in this paper. Compared with re- gion relationship tree proposed by Fub, global contour of image can be described in our methods and it made the relationship of contours more accurately and integrally. The methods can also be used in applications of image representation, such as image recon- structing and image matching. Experiments show that it have excellent performance in image representation.
出处
《微计算机信息》
北大核心
2008年第21期274-276,共3页
Control & Automation
基金
国家863计划资助项目(2006AA12A104)
关键词
图像表达
轮廓层次
局部轮廓
图像匹配
Image representation
contour hierarchica
local contour
image matching
