Generation of reference data of 3D surface texture using the non-causal 2D AR model

In this paper, the non-causal 2D AR model was proposed as a new method for generation of reference data of the 3D surface texture. Specification parameters to engender various topographical properties were the correlation distances in the x and y directions, the power index which determined the decaying pattern of the auto-correlation function, the root mean deviation, the skewness and the kurtosis. Isotropic surfaces were generated and compared with results by the causal 2D AR model which has been proposed in the past researches. It was found that the error of generation by the non-causal 2D AR model was smaller than that by the causal 2D AR model. Anisotropic surfaces and non-Gaussian distributed surfaces were also generated successfully. (C) 2004 Elsevier B.V. All rights reserved.