基于特殊点跟踪的线状目标矢量化研究  

Vectorizing research of linear target based on special points tracking

作　　者：彭靖叶[1] 孙永荣[1] 黄斌[1] 

机构地区：[1]南京航空航天大学自动化学院,南京210016

出　　处：《计算机工程与应用》2015年第15期198-201,共4页Computer Engineering and Applications

摘　　要：针对单像素线宽线状目标栅格图像矢量化,提出了一种基于特殊点跟踪的矢量化算法。该算法通过提取曲线端点和曲线间交叉点这两类特殊点,以扫描跟踪的方式提取曲线上的所有像素点坐标,采用Douglas-Peucker法对曲线栅格点进行压缩,获得最终矢量图。实验结果表明,该算法能够有效地实现线状目标栅格图像的矢量化,在选取数据压缩阀值为2个像素的情况下,矢量图与原栅格图的距离差均值小于0.4个像素,数据压缩比大于850。An automatic vectorizing algorithm based on special-point tracking is developed for the raster image of one-pixelwidth linear target. The end points and crossing points of the curve are picked up, which are then used to extract all the pixel points in the curves. The Douglas-Peucker method compresses the pixel points to obtain the final vector image. The test results show that the algorithm is available and valid. The mean distance between the vector image and the raster image is less than 0.4 pixels, and the compression ratio is greater than 850, when the compression threshold is 2 pixels.

关 键 词：图像矢量化 线状目标 特殊点跟踪 曲线压缩 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]