基于颜色二阶矩和直方图联合判定的镝铁中铁含量自动分析仪研制

镝铁合金中铁含量的检测方法通常以人工滴定为主，但这种方法存在耗时长、操作繁琐、人眼判断滴定终点差异性等问题。本文研制了一种以图像处理为基础的镝铁中铁含量自动分析仪，实现了样品前处理、滴定、终点判定等过程一体化与自动化，减少了人工操作。可同时进行五个样品试验，提高了滴定效率。采用CCD工业相机摄取溶液图像，实现化学反应过程中溶液颜色的自动化识别，运用颜色分量二阶矩阈值和像素点数量排除溶液颜色交替变化的干扰，并基于直方图对比得到相似度进行滴定终点的判定，能够有效、准确地实现化学反应过程中溶液颜色的自动化识别。使用参考值为19.47%-20.25%的镝铁合金样品进行试验，与手工滴定结果比对验证，精密度测量结果相对标准差在0.21%以内，正确度测量结果误差在0.15%以内，满足国标《镝铁合金化学分析方法第4部分：铁量的测定 重铬酸钾容量法》（GB/T26416.4—2010）的精度要求。

Development of an automatic analyzer for iron content in dysprosium iron based on the joint determination of color second-order moment and histogram

The detection method of iron content in dysprosium ferroalloy is usually manual titration, but this method has some problems, such as time-consuming, cumbersome operation, human eyes judging the difference of titration end point and so on. In this paper, an automatic analyzer for iron content in dysprosium iron based on image processing is developed, which realizes the integration and automation of sample pretreatment, titration and end point determination, and reduces manual operation. Five samples can be tested at the same time, which improves the titration efficiency. The CCD industrial camera is used to capture the solution image to realize the automatic recognition of solution color in the process of chemical reaction. The dysprosium ferroalloy samples with reference value of 19.47% - 20.25% are used for the test, and compared with the manual titration results. The relative standard deviation of the accuracy measurement results is within 0.21%, and the error of the accuracy measurement results is within 0.15%, which meets the accuracy requirements of the national standard methods for chemical analysis of dysprosium ferroalloy Part 4: Determination of iron content potassium dichromate volumetric method (GB / t26416.4-2010).