一种新的图件扫描影像细化算法

针对传统的细化算法如经典细化算法、跟踪剥皮算法等在细化时需花费大量的时间这一问题，提出了一种基于最短距离边缘跟踪技术的细化算法，此算法通过对线宽两边相应边缘点中点的求取来细化并同时矢量化，能有效地提高细化速度。

图件扫描数字化的误差分析

从图件扫描数字化误差理论分析出发，利用定向点和采样点的自动对中和多种人工采点方法的实验数据，对误差原因及分布规律进行了验证，得出了扫描数字化精度与手扶跟踪数字化精度相比具有一定的优势，对提高大比例尺图件数字化的精度和速度具有借鉴意义．

扫描仪的校准与图件扫描纠正

原有图件的数字化过程中,扫描仪器的校准和图件的扫描纠正对数据采集的精度来说至关重要,作者结合实际工作经验作了一些简要的总结。

一个扫描图件的图文分离算法

在分析传统图文分离算法的基础上 ,提出了一种新的算法。它由图像连通区域宽度和高度频数直方图 ,求出字符的最大 (小 )宽度 (高度 ) ,从而较准确地完成图文分割 。

大比例尺化探数据库建库工作流程

通过对大比例尺化探数据库建库的试点工作 ,总结出一套数据采集的工作方法 。

Calculation of Percentage of Coarse Aggregate Present in Concrete Using Processing of Digital Images Obtained with a Commercial Scanner

This paper presents a method for determining the percentage of coarse aggregate in concrete specimens by image processing. The test pieces were produced with the aim of obtaining images of their cross sections through a scanner table. In order to increase the contrast between mortar and coarse aggregate the sliced surfaces were treated with the phenolphthale in solution. The images obtained in the scanner were processed in a program developed with MATLAB （matrix laboratory）. The average coarse aggregate in each section and the mean of coarse aggregate per test body were calculated. With the results, it was revealed that the method returned satisfying results when compared to the original trace of the concrete.