Abstract
Data compression by using the Discrete Wavelet Transform has a problem in that the errors caused by irreversible process can be obtained only by comparing the original data with the data after compression and expansion. Such a back-and-fourth process is not efficient especially for huge Zmap models of GB sizes. We propose an algorithm for limited error compression for CAD models. By using our method, the user can specify the maximum error before compressing the shape data,which enables to preserve the whole shapes through NC milling simulation.
