The iron ore sinter quality is described by various parameters. The mineralogical and textural characteristics are obtained by image processing from microimages of polished sections. The image processing software VisuMet is adapted to analyze the mineral abundance and porosity of iron ore sinter. New algorithms are developed to calculate microstructural characteristics, like the pore size distribution and the effective thickness of sinter grains. The validation of the method using image processing is given by the description of three test series to estimate the systematic and analytical errors. The application to five industrial sinter samples is outlined. Ninety percent of all pores are smaller than 250 μm in equivalent diameter and the mean equivalent diameter ranges from 76 to 79 μm. The effective thickness, which is defined as the maximum distance between solid and nonsolid sinter material, varies between 120 and 368 μm. The area degradation algorithm of VisuMet is used to calculate the gradient between 30% and 80% of the degraded area. Dense sinter grains have a lower gradient than porous ones.

中文翻译：

---

图像处理对铁矿石烧结矿的扩展分析

铁矿石烧结矿的质量由各种参数来描述。矿物学和质地特征是通过对抛光部分的显微图像进行图像处理而获得的。图像处理软件VisuMet适用于分析铁矿石烧结矿的矿物丰度和孔隙率。开发了新的算法来计算微观结构特征，例如孔径分布和烧结晶粒的有效厚度。通过三个测试系列的描述来评估使用图像处理方法的有效性，以评估系统误差和分析误差。概述了在五个工业烧结矿样品中的应用。所有孔的百分之九十的等效直径小于250μm，平均等效直径范围为76至79μm。有效厚度 固态和非固态烧结材料之间的最大距离定义为120至368μm。VisuMet的面积退化算法用于计算退化面积的30％至80％之间的梯度。致密的烧结晶粒比多孔的晶粒具有较低的梯度。