Adaptive denoising method to improve aberration measurement performance

Abstract

An adaptive denoising method utilizing weighted least-square (WLSQ) is proposed for an aberration measurement (AM) method based on aerial images (AI). In the aberration measurement process, the WLSQ method is employed to obtain the principal component coefficients which are used for extracting the Zernike coefficients. In this paper, we also provide a noise model according to the actual aerial images. Based on this noise model, a simplified weighting factor for the WLSQ method is also proposed. Because of the adaptive and lossless denoising ability of the proposed method, more accurate Zernike coefficients can be extracted from aerial images. Compared with the aberration measurement techniques based on principal component analysis of aerial images (AMAI-PCA), the proposed method can enhance the accuracy by more than 30% when the range of Zernike aberrations is within 0.1λ. The experiment also shows that the proposed method can detect the aberration shift more accurately.