改进的Potrace提花织物图像矢量化算法  

Improved Potrace algorithm for vectorization of image of jacquard

作　　者：姚鹏鹏[1] 樊臻[1] 张森林[1] 

机构地区：[1]浙江大学电气工程学院,浙江杭州310007

出　　处：《传感器与微系统》2014年第4期125-127,共3页Transducer and Microsystem Technologies

基　　金：国家科技支撑计划资助项目(2013BAH58F02)

摘　　要：为了对提花织物图像进行矢量化,针对其颜色少、色块大的特点,提出了改进的Potrace图像矢量化算法。原始的Potrace算法只能实现对二值图像的矢量化,改进后的算法将位图中的色块逐个分解生成一个个的闭合路径,之后将这些闭合路径按照其各自分布拼接成树状结构并矢量化,最终生成一个完整的矢量图形。该算法在实际的应用中取得较好效果。In order to vectorize jacquard image,aiming at small amount of color and large color blocks,propose an improved Potrace algorithm for vectorization of image.The original Potrace algorithm can only realize vectorization of binary image and the improved algorithm decompose color blocks of bitmap one by one to closed paths which are combined to a tree structure later and merged to a vector graphic at last.This algorithm achieves good effect in practical use.

关 键 词：提花织物 Potrace算法 图像矢量化 路径分解 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]