This paper presents methods to calculate the length overestimation errors which are being made when approximating a discrete line by edges of triangular tessellation and marching squares algorithm. The maximum error and its average value are 15.47 % and 10.27 % for the triangular tessellation, while for the marching squares approach they are 8.24 % and 5.49 %, respectively. Mathematical calculations were compared with experimental results obtained by the Electron Backscatter Diffraction technique showing their usefulness as correction coefficients to obtain more accurate boundary length estimates.

中文翻译：

---

三角细分和行进平方算法估计离散线逼近的系统误差

本文提出了一些方法来计算长度高估误差，这些误差是通过三角细分和行进平方算法逼近离散线而产生的。三角形细分的最大误差及其平均值为15.47％和10.27％，而行进方差法则分别为8.24％和5.49％。将数学计算与通过电子背向散射衍射技术获得的实验结果进行了比较，结果表明它们可用作校正系数以获得更准确的边界长度估计的有用性。