基于机器视觉的短纤维复合材料的取向度提取方法  

Extracting orientation index of short fiber reinforced composites by computer vision methods

作　　者：郑子君 乔英 邵家儒 ZHENG Zijun;QIAO Ying;SHAO Jiaru(Department of Mechanical Engineering,Chongqing University of Technology,Chongqing 400054,China)

机构地区：[1]重庆理工大学机械工程学院,重庆400054

出　　处：《复合材料科学与工程》2024年第3期65-72,共8页Composites Science and Engineering

基　　金：国家自然科学基金(11702046);重庆市教委科学技术研究项目(KJQN202201113,KJQN202201105)。

摘　　要：纤维取向度对短纤维增强复合材料的宏观性质有着明显的影响。采用机器视觉方法从扫描电镜(SEM)照片中提取纤维取向度是一种兼顾效益和成本的做法。为准备训练和测试数据集,基于正交椭圆纤维闭合近似模型推导了一种取向分布,并结合接收-拒绝算法和RSA算法实现了短纤维复合材料几何结构的重构,从而生成大量模拟SEM图像。在此基础上,提出一种基于灰度共生矩阵的BP神经网络(GLCM-BP)用于取向度的预测,并与常见的形态学分割、结构张量关系、卷积神经网络算法(CNN)方案的结果进行了对比。算例表明,GLCM-BP模型能够很好地预测纤维取向度,测试数据的拟合相关性达到0.99,均方误差约为0.01,满足工程使用的需要。横向对比发现:结构张量公式在平面分布中的预测结果明显偏小;在低纤维体积分数时,形态学方法和GLCM-BP法预测结果最好;在高纤维体积分数时,GLCM-BP和卷积神经网络表现更好。本文提出的GLCM-BP法也有一定抵抗噪点的能力。The orientation of fibers has a significant impact on the macroscopic properties of short fiber rein-forced composite materials.The fiber orientation index is extracted from scanning electron microscope(SEM)images by machine vision methods.To build the training/testing data sets,a fiber orientation distribution is derived based on the orthogonal elliptic closed approximation,and then many simulated SEM images are generated by using the acceptance-rejection and random sequential adsorption algorithms.Based on these simulated images,a BP neu-ral network based on gray-level co-occurrence matrix(GLCM-BP)was proposed to predict the fiber orientation in-dex,and the results were compared with commonly used methods,including morphological segmentation,structure tensor relationship,and convolutional neural network(CNN)algorithms.The results showed that the GLCM-BP model could effectively predict the fiber orientation with a fitting correlation of 0.99 and a mean square error of ap-proximately 0.01,meeting engineering requirements.In comparison,the structure tensor formula systematically un-derestimates the orientation in planar distributions;morphological and GLCM-BP methods perform better for low fi-ber volume fractions;GLCM-BP and CNN methods perform better for high fiber volume fractions.The proposed GL-CM-BP method also shows capability to resist image noise.

关 键 词：短纤维增强复合材料 取向度 人工神经网络 形态学 机器视觉 

分 类 号：TB332[一般工业技术—材料科学与工程]