A parallel thinning method based on image marking

Abstract

本文提出了一种基于图像标记的并行细化算法 .首先计算图像中像素和轮廓距离的标记信息 ,并将标记图像作为细化算法的输入 .通过综合利用标记信息 ,保留了几何特性 ,提高了细化算法的鲁棒性 ,同时为进一步图元识别 /拟合奠定了基础. A new parallel thinning method based on image marking is proposed in this paper.First,the distance from every pixel to its nearest contour is calculated.Then these distances will be constructed as a marking image of the original one and inputted for thinning.The experimental results have demonstrated that the proposed approach can achieve comparable satisfactory performance in terms of keeping geometrical properties and robustness to single pixel noise by the integrative usage of the marking information.Moreover,the method presented will also benefit the basis for further graph recognition.