| 引用本文: | 刘庆生,曾少军,卢小能,宋翰林.基于CT图像的铝电解阴极炭块组分识别研究[J].材料科学与工艺,2017,25(4):63-70.DOI:10.11951/j.issn.1005-0299.20160223. |
| LIU Qingsheng,ZENG Shaojun,LU Xiaoneng,SONG Hanlin.Study on CT image based composition identification of aluminum electrolytic cathode blocks[J].Materials Science and Technology,2017,25(4):63-70.DOI:10.11951/j.issn.1005-0299.20160223. |
|
| 摘要: |
| 本文为铝电解用阴极炭块检测提供一种科学的方法，通过对铝电解用阴极炭块中炭骨料成分的颗粒级配计算，由骨料颗粒个数比例、面积比例分析铝电解用阴极炭块的质量.通过分析铝电解阴极炭块显微CT数字图像，基于数学形态学研发了一种实现其组分定量检测的技术方法.利用显微CT设备获取阴极炭块样品的数字图像及灰度值矩阵，基于显微CT成像原理区别出阴极炭块的组成成分为空隙、沥青、炭骨料及杂质，通过对阴极炭块数字图像校正光密度、校正颜色、选择阀值、填充空隙、检测边缘、分离区域及进行腐蚀、膨胀、开运算与闭运算等其他数学形态学特征计算.实验得到样品实际组成与处理结果数值处于合理的偏差范围内，得出显微CT数字图像处理是研究铝电解阴极碳块组分识别及特征提取的简便可靠的方法.结果表明，基于CT断层扫描图片可以较准确地实现阴极炭块组成成分的虚拟识别.
|
| 关键词: 铝电解 显微CT 图像处理 阴极炭块 粒径级配 |
| DOI：10.11951/j.issn.1005-0299.20160223 |
| 分类号:TF821 |
| 文献标识码:A |
| 基金项目:国家自然科学基金资助项目(1,9). |
|
| Study on CT image based composition identification of aluminum electrolytic cathode blocks |
|
LIU Qingsheng,ZENG Shaojun,LU Xiaoneng,SONG Hanlin
|
|
(School of Metallurgy and Chemical Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China )
|
| Abstract: |
| In this paper, the aluminum electrolysis cathode carbon block detection provides a scientific method to directly and efficiently calculate the particle size distribution of carbon aggregate composition in the cathode carbon block of the aluminum electrolysis. The quality of cathode carbon block for aluminum electrolysis can be obtained by particle area ratio and particle number proportion.Through the micro CT digital image analysis of the cathode carbon block, a quantitative component detection method was proposed by mathematical morphology. Digital image and gray matrix of the cathode carbon block can be obtained by using the micro CT. The voids, asphalt, aggregate and impurities can be distinguished by micro CT imaging, including, optical density correction, color correction, threshold selection, void filling, edge detection, isolation and expansion of regional corrosion, opening operation and closing operation of mathematical morphology and other characteristics of calculation. The results obtained here are identical to the actual chemical composition within a reasonable deviation range, indicating that the micro CT digital image processing is a simple and reliable method of component identification and feature extraction of aluminum electrolytic cathode carbon block. The verification results show that, based on CT tomography images can accurately realize the virtual identification of the components of the cathode carbon blocks.
|
| Key words: aluminum electrolytic micro CT image processing cathode carbon block particle size distribution |
