CAJ | 학술논문

能否准确对精密光学元件表面评价势必影响着光学元件研制与生产，其中中频误差的提取逐渐显得重要。提出了利用双树复小波变换( DT－CWT)对精密光学元件表面提取信号进行多尺度分解，以定义的均方根波长确定其双树复小波分解次数，进行中频误差的提取与识别。此方法应用于精磨和镀膜的精密光学元件表面中，实验证实该方法不仅可以应用于一维信号的提取，同样也适合三维表面中频波段面形误差的提取。
능부준학대정밀광학원건표면평개세필영향착광학원건연제여생산，기중중빈오차적제취축점현득중요。제출료이용쌍수복소파변환( DT－CWT)대정밀광학원건표면제취신호진행다척도분해，이정의적균방근파장학정기쌍수복소파분해차수，진행중빈오차적제취여식별。차방법응용우정마화도막적정밀광학원건표면중，실험증실해방법불부가이응용우일유신호적제취，동양야괄합삼유표면중빈파단면형오차적제취。
The ability to evaluate the precision optical component surface accurately is bound to affect the development and produc-tion of optical components, more and more attention is paid to researching on the intermediate-frequency surface shape error. This paper uses the dual tree complex wavelet transform ( DT -CWT) to proceed a multi-scale decomposition of a signal extracted from the precision optical component surface, applies the defined root-mean-square wavelength to determining the number of dual tree complex wavelet decomposition. lt conduct the intermediate-frequency error extraction and recognition. This method is applied to fine-grinding and coating precision optical component surfaces, and the experiment proves this method not only can be applied to one di-mensional signal extraction, but also suitable for the extraction of 3-D medium frequency band surface shape error.