用于轴承表面缺陷分类的特征选择算法

Feature Selection Algorithm for Classification of Bearing Surface Defects

针对现有基于机器视觉的轴承表面缺陷类型识别中分类特征选择这一薄弱环节,提出一种综合运用相关分析、标量特征选择和特征向量选择的实用特征选择算法。首先,通过相关分析剔除相似特征;然后,用标量特征选择算法作进一步筛选;最后,用特征向量选择算法选出最终分类特征。对比试验表明:该算法可实现有效的特征选择,识别率高达99.5%,且避免了大规模运算。

Aiming at shortcoming of classification feature selection in type recognition of bearing surface defects based on machine vision,a practical feature selection algorithm is proposed based on correlation analysis,scalar feature selection and feature vector selection. Firstly,the similar features are removed by correlation analysis. Then,the scalar feature selection algorithm is used for further screening. Finally,the final classification feature is selected by feature vector selection algorithm. The comparative experiments show that the algorithm achieve effective feature selection with a recognition rate as high as 99. 5 % and avoid large-scale computation.