基于PCA的零件表面质量检测新方法

The Application of PCA in the Feature Extraction of Digital Figure

介绍了在小波变换的基础上运用主成份分析法对铸造零件表面缺陷数字图像进行特征提取,提出了简化零件表面质量自动检测计算量的新方法,具体地阐述了主成份分析法的原理、计算方法、数字图像分割、特征提取,并通过实例分析进行优化参数选取,具有实际应用价值。

In this thesis, the digital figure of the defect of the casting elements is extracced by PCA. And a new method that simplifies the automatic quality detection is also brought forward. The author not only expounds the theory of PCA,the calculating method,the segmatatian of digital figure and the feature extration, but also analysze the real example to optimize the selecting of the preferences. This thesis is valuable in the real manufacture.