摘要
骨架保存了要处理对象的拓扑信息,是图像分析的重要研究内容之一.传统的骨架细化算法不能保证结果的准确性,而距离场的方法无法保证结果的连续性.为此提出一种快速有效的骨架提取算法,将经典的距离变换法和细化方法结合,克服二者之间存在的缺陷,实现算法的互补.经过大量实验验证,此方法能够得到连续、准确的骨架,可以很好地满足实际应用的需求.
Skeleton contains the topological information of object,which is an important search aspect in image analysis.Traditional thinning algorithm for skeleton extraction can＇t guarantee the accuracy,while distance field algorithm can＇t guarantee continuity.A new method for skeleton extraction is proposed,which combines the distance transformation and thinning algorithm so as to overcome the disadvantage of the two algorithms.Experimental results shows that proposed method can extract accurate and continuous skeletons,which can satisfy the demand for practical application.
出处
《微电子学与计算机》
CSCD
北大核心
2011年第10期114-117,121,共5页
Microelectronics & Computer
基金
国家自然科学基金项目(60775015)
江苏省青蓝工程
关键词
距离场
骨架提取
细化
distance field
skeleton extraction
thinning
