Assessment of importance-based machine learning feature selection methods for aggregate size distribution measurement in a 3D binocular vision system
Assessment of importance-based machine learning feature selection methods for aggregate size distribution measurement in a 3D binocular vision system
复制标题
评估基于重要性的机器学习特征选择方法，用于 3D 双目视觉系统中聚合尺寸分布测量
DOI：
10.1016/j.conbuildmat.2021.124894
复制
发表时间：
2021
影响因子：
7.4
通讯作者：
Lili Pei
中科院分区：
文献类型：
--
作者：
Zhaoyun Sun;Hanye Liu;Ju Huyan;Wei Li;Meng Guo;Xueli Hao;Lili Pei
骨料粒度通常通过人工取样和筛分来测量。机器视觉技术可以提供快速、非侵入性的测量。然而，传统的成像方法使用一个单一的尺寸描述符来区分不同的筛径级的粗骨料可能不会产生高精度的分类结果。为了确定粗集料筛分测量的最佳监督机器学习模型，对17种方法进行了评估和比较。为了训练我们的模型，引入了一个名为MFCA 27（Multiple Features of Coarse Aggregate 27）的新数据集，该数据集包含基于聚合三维（3D）顶面对象的27个聚合特征。此外，开发了一种用于调查不同特征集下的准确性如何随数据集而变化的特征选择方法，其中根据使用极端随机化树模型测量的基于杂质的特征重要性得分来执行特征选择。实验表明，高斯过程分类器（GPC）是最好的性能与二维或三维（2D/3D）的特征集的数据集上的方法，在准确性和鲁棒性。与传统的基于单一粒径描述符的团聚体粒径测量方法相比，GPC在MFCA 27测试数据集上的团聚体粒径等级测量准确率可达95.06%。
Aggregate size is usually measured by manual sampling and sieving. Machine vision techniques can provide fast, non-invasive measurement. However, the traditional imaging method using a single size descriptor to discriminate different sieve-size classes of coarse aggregates might not yield high-precision classification results. To determine the optimum supervised machine learning model for coarse aggregates sieve-size measurement, 17 methods were evaluated and compared. To train our model, a new dataset named MFCA27 (Multiple Features of Coarse Aggregate 27) was introduced, which contains 27 features of aggregates based on aggregate three-dimensional (3D) top-surface object. In addition, a feature selection approach for investigating how accuracy varied with the datasets under different feature sets was developed, where feature selection was performed according to the impurity-based feature importance score measured using an extremely randomized tree model. Experiments demonstrated that the Gaussian process classifier (GPC) was the best-performing method on the datasets with two- or three-dimensional (2D/3D) feature sets in terms of accuracy and robustness. The results also showed that, compared with the traditional aggregate sieve-size measurement method, which is based on a single size descriptor, GPC can achieve an accuracy of 95.06% on the test dataset of MFCA27 in the aggregate sieve-size class measurement task.