摘要
提出了一种基于伪Zernike矩不变量分析的视觉测力方法.该方法利用伪Zernike矩不变量所构成的特征向量来描述微装配过程中微夹爪变形之后的形状,并由此建立了矩不变量特征向量与受力之间的函数关系;建立训练集,输入为矩不变量特征向量,输出为已知受力;通过支持向量机比较测试集与训练集中的特征向量,对测试集的输入进行多类分类,从而估计未知受力.对4种不同规格的微悬臂梁进行了实验,结果验证了该方法的有效性.
A new visual force measurement method was proposed to measure applied forces through analyzing pseudo Zernike moment invariants. First, a feature vector of pseudo Zernike moment invariants is used to describe the deformed microgripper shape in microassembly. This description indicates a function between the applied force and the feature vector of moment invariants. Then, a training set is established, where the input and the output represent the feature vector of moment invariants and the corresponding known applied force, respectively. Finally, a support vector machine (SVM) compares the feature vectors in the testing set with those ones in the training set, implements multi-class classification and estimates unknown applied forces in the testing set. The experiments on four different microcantilevers validate the proposed method.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2013年第4期589-593,共5页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金资助项目(50875169)
关键词
视觉测力
伪Zernike矩不变量
支持向量机
多类分类
visual force measurement
pseudo Zernike moment invariants
support vector machine
multiclass classification