摘要
根据参数曲线集区域识别算法的自动化和通用性要求,采用多叉树形式化表达区域的空间组织结构,在建立定向闭环的基础上,提出复杂区域(内含孔、岛)的全自动识别方法。与其他区域识别算法相比,该算法对边界粘连及内含孔、岛的区域均能有效地识别与提取,具有广泛的适应性且易于实现。
According to the automation and general requirements of parameterized curve set regions identification algorithm,spatial organization of regions generated by parameterized curve set is represented using multi-tree structure.Based on the establishment of directional closed loops,a new method for identification of complex regions is proposed for the automation and general applicability.Compared with the other methods,this method is effective for regions with holes and islands.It has wide adaptability and is easy to implement.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第8期23-26,共4页
Computer Engineering
基金
国家自然科学基金资助项目(50774092)
关键词
参数曲线
区域识别
多叉树
定向闭环
parameterized curve
region identification
multi-tree
directional closed loop
