- 學位論文
反向量化方法應用於遞迴傅立葉演算法的研究
The Study of the Iterative Quantization Methods for the Iterative Fourier-Transform Algorithm
摘要
以二元光學的方式來製作繞射光學元件，必須將連續相位分布的繞射元件作相位量化，然而此過程將無可避免的產生量化誤差。雖然增加量化階數可以有效地提高繞射效率，不過製作困難度也相對地提高。因此，使用低階非均勻量化之繞射光學元件為相對可行之方法。 在此次模擬設計中，我們將模擬出四種反向量化方法(backward iterative quantization)來設計多階繞射光學元件，來用以比較較廣泛地被使用地量化方法(forward iterative quantization)，也用來比較四階繞射光學元件中均勻與非均勻量化相位之結果，我們發現反向遞迴量化的確可以提高繞射效率。再者，第四種量化方法可較達到較高的效率也可較達到較高的訊雜比。然而，我們發現反向量化所得到的結果遠較傳統的量化方法穩定且效率較高，繞射效率皆在大約在82%左右。
並列摘要
It is essential to quantize the phase of continuous-phase diffractive optical elements (DOEs) when they are implemented using the binary optics. However, the phase quantization introduces the quantization error. Although the use of higher number of quantization phases effectively increases diffraction efficiency, the difficulty of manufacturing DOEs increases. The DOEs with unequally spaced quantization phases can achieve high diffraction efficiency with the smaller number of phases. In simulations, we used four types of backward iterative quantization methods for designing multilevel DOEs. The widely-used iterative quantization method was compared with the proposed methods in the DOE designs with equally and unequally spaced phases. It was observed that the fourth method can achieve higher diffractive efficiency and signal-to-noise ratio. Nevertheless, we found that the backward methods were more stable than traditional quantization method, and achieved higher efficiency(approximately 82%). In this, thesis, different quantization methods were discussed and compared to illustrate their influences on the unequal-phase DOEs.