论文信息 - Consequences of stratified sampling in graphics

Consequences of stratified sampling in graphics

central limit theorem predicts a variance of the mean of O(N -1). Stratified sampling can further reduce the variance of the me This paper investigates how and why stratification effects t convergence to mean value of image pixels, which are obser to converge from N-2 to N-1, with a rate of about N -3/2 in pixels containing edges. This is consistent with results from the the of discrepancy. The result is generalized to higher dimensio as encountered with distributed ray tracing or form-fact computation.

Don P. Mitchell | D. P. Mitchell

[1] R. R. Coveyou. Monte Carlo Principles and Neutron Transport Problems , 1971 .

[2] J. Halton. A Retrospective and Prospective Survey of the Monte Carlo Method , 1970 .

[3] Samuel P. Uselton,et al. Statistically optimized sampling for distributed ray tracing , 1985, SIGGRAPH.

[4] J. Hammersley,et al. Monte Carlo Methods , 1965 .

[5] David Eppstein,et al. Computing the discrepancy with applications to supersampling patterns , 1996, TOGS.

[6] Don P. Mitchell,et al. Generating antialiased images at low sampling densities , 1987, SIGGRAPH.

[7] Robert L. Cook,et al. Stochastic sampling in computer graphics , 1988, TOGS.

[8] James T. Kajiya,et al. The rendering equation , 1986, SIGGRAPH.

[9] Mark A. Z. Dippé,et al. Antialiasing through stochastic sampling , 1985, SIGGRAPH.

[10] J. Beck,et al. Irregularities of distribution , 1987 .

[11] Peter Shirley,et al. Discrepancy as a Quality Measure for Sample Distributions , 1991, Eurographics.

[12] J. Painter,et al. Antialiased ray tracing by adaptive progressive refinement , 1989, SIGGRAPH.

[13] David P. Dobkin,et al. Random-edge discrepancy of supersampling patterns , 1993 .