形面测量误差频域阈值处理方法  

Frequency domain threshold method for profile measurement error processing

作　　者：吕东方[1] 张立富[1] 刘奇晗[2] 

机构地区：[1]黑龙江大学机电工程学院,哈尔滨150080 [2]哈尔滨工业大学媒体技术与艺术系,哈尔滨150001

出　　处：《黑龙江大学工程学报》2011年第4期113-117,共5页Journal of Engineering of Heilongjiang University

基　　金：黑龙江省教育厅科学技术研究项目(11551331)

摘　　要：采用基于小波变换的频域方法对轮廓测量数据中的误差进行处理。阐述了小波阈值方法处理数据误差的基本原理,选取不同的阈值分别对测量数据中的粗大误差点和随机误差点进行处理。通过分析形面测量数据小波高频分解系数的特征,应用3σ准则对数据点中的粗大误差数据进行剔除,而后采用基于D.L Donoho阈值方法的层进阈值函数对测量数据中的随机误差点来进行处理。通过系统的比对具有代表性的小波函数阈值误差处理效果,结果表明采用的阈值去噪方法优于传统的阈值方法。The wavelet threshold de-noising algorithm is applied to process errors in the profile measurement data. The error processing priciple based on the wavelet tranform is represented in the paper, and the different threshold methods are selected for the gross errors and the random errors. According to the data character in the high frequency coefficients of the wave-let transform, the 3σ criterion is used for the gross errors deletion and the layer function based on D. L. Donoho is applied to reduce the random errors. The processing errors based on the canonical wave-let functions were systematically compared and it was concluded that the method in the paper could effectively decrease the profile measurement error.

关 键 词：误差处理 小波变换 阈值方法 

分 类 号：TH701[机械工程—仪器科学与技术]