Answers are grounded in this paper and cite their sources.
Fourier transforms have been successfully employed to produce fast, accurate calculations for dose precompensation in proximity effect correction for electron beam lithography. The principle advantage of this method is that it lends itself to implementation on special purpose computing hardware which is optimized for the generation of one- and two-dimensional Fourier transforms. However, the extremely accurate data base produced by the direct employment of the method is excessively large and, therefore, not practical. A controlled method for approximating this accurate solution is required which gives rise to the least mean-square error in the net exposure. This process of approximation can reduce the data base by ‘‘thinning’’ or selectively removing larger and larger features. In this paper we explore the use of two-dimensional Haar transforms for thinning the data base. An additional benefit of using Haar transforms is that they may also be computed with the same hardware accelerators which speed up the calculation of Fourier transforms.
No takes yet. Share an insight, caveat, or question.
Haslam et al. (1985) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: