摘要
针对一般细化算法在进行图像细化时会出现细化不彻底、过度腐蚀等问题,提出一种改进的A-W细化算法。改进算法采用高斯平滑滤波器对原始图像进行平滑,在A-W细化算法的20种细化模板基础上,增加3组细化模板进行第1次细化,并设计4种新的细化模板进行第2次细化。实验结果证明,改进算法能较好保持曲线的连通性,保留主线的细节特征,细化后的图像光滑无毛刺。
Because the general thining algorithms often generate problems such as multiple-pixel width and excessive corrosion, this paper proposes an improved rotation invariant thinning algorithm named A-W algorithm. The original image is smoothed by Gaussian smoothing filter, and then is thinned the first time by using the A-W thining method which adds three groups of thining templates and four new kinds of thining templates are designed to achieve the second thining. Experimental results show that the proved algorithm can better keep connectivity of the curve, keep the details of the main features, and after thinning of the image smooth without burr.
出处
《计算机工程》
CAS
CSCD
2012年第6期219-220,223,共3页
Computer Engineering
关键词
细化算法
A-W算法
高斯平滑滤波器
单像素
细化模板
thining algorithm
A-W algorithm
Gaussian smoothing filter
single pixel
thining template
