Sampling is important for many applications in research areas such as graphics, vision, and image processing. In this paper, we present a novel stratified sampling algorithm (SSA) for the coiled tubing surface with a given probability density function. The algorithm is developed from the inverse function of the integration for the areas of the coiled tubing surface. We exploit a Hierarchical Allocation Strategy (HAS) to preserve sample stratification when generating any desirable sample numbers. This permits us to reduce variances when applying our algorithm to Monte Carlo Direct Lighting for realistic image generation. We accelerate the sampling process using a segmentation technique in the integration domain. Our algorithm thus runs 324 orders of magnitude faster when using faster SSA algorithm where the order of the magnitude is proportional to the sample numbers. Finally, we employ a parabolic interpolation technique to decrease the average errors occurred for using the segmentation technique. This permits us to produce nearly constant average errors, independent of the sample numbers. The proposed algorithm is novel, efficient in computing and feasible for realistic image generation using Monte Carlo method.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Chung-Ming WANG, Peng-Cheng WANG, "A Novel Approach to Sampling the Coiled Tubing Surface with an Application for Monte Carlo Direct Lighting" in IEICE TRANSACTIONS on Information,
vol. E87-D, no. 6, pp. 1545-1553, June 2004, doi: .
Abstract: Sampling is important for many applications in research areas such as graphics, vision, and image processing. In this paper, we present a novel stratified sampling algorithm (SSA) for the coiled tubing surface with a given probability density function. The algorithm is developed from the inverse function of the integration for the areas of the coiled tubing surface. We exploit a Hierarchical Allocation Strategy (HAS) to preserve sample stratification when generating any desirable sample numbers. This permits us to reduce variances when applying our algorithm to Monte Carlo Direct Lighting for realistic image generation. We accelerate the sampling process using a segmentation technique in the integration domain. Our algorithm thus runs 324 orders of magnitude faster when using faster SSA algorithm where the order of the magnitude is proportional to the sample numbers. Finally, we employ a parabolic interpolation technique to decrease the average errors occurred for using the segmentation technique. This permits us to produce nearly constant average errors, independent of the sample numbers. The proposed algorithm is novel, efficient in computing and feasible for realistic image generation using Monte Carlo method.
URL: https://global.ieice.org/en_transactions/information/10.1587/e87-d_6_1545/_p
Copy
@ARTICLE{e87-d_6_1545,
author={Chung-Ming WANG, Peng-Cheng WANG, },
journal={IEICE TRANSACTIONS on Information},
title={A Novel Approach to Sampling the Coiled Tubing Surface with an Application for Monte Carlo Direct Lighting},
year={2004},
volume={E87-D},
number={6},
pages={1545-1553},
abstract={Sampling is important for many applications in research areas such as graphics, vision, and image processing. In this paper, we present a novel stratified sampling algorithm (SSA) for the coiled tubing surface with a given probability density function. The algorithm is developed from the inverse function of the integration for the areas of the coiled tubing surface. We exploit a Hierarchical Allocation Strategy (HAS) to preserve sample stratification when generating any desirable sample numbers. This permits us to reduce variances when applying our algorithm to Monte Carlo Direct Lighting for realistic image generation. We accelerate the sampling process using a segmentation technique in the integration domain. Our algorithm thus runs 324 orders of magnitude faster when using faster SSA algorithm where the order of the magnitude is proportional to the sample numbers. Finally, we employ a parabolic interpolation technique to decrease the average errors occurred for using the segmentation technique. This permits us to produce nearly constant average errors, independent of the sample numbers. The proposed algorithm is novel, efficient in computing and feasible for realistic image generation using Monte Carlo method.},
keywords={},
doi={},
ISSN={},
month={June},}
Copy
TY - JOUR
TI - A Novel Approach to Sampling the Coiled Tubing Surface with an Application for Monte Carlo Direct Lighting
T2 - IEICE TRANSACTIONS on Information
SP - 1545
EP - 1553
AU - Chung-Ming WANG
AU - Peng-Cheng WANG
PY - 2004
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E87-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 2004
AB - Sampling is important for many applications in research areas such as graphics, vision, and image processing. In this paper, we present a novel stratified sampling algorithm (SSA) for the coiled tubing surface with a given probability density function. The algorithm is developed from the inverse function of the integration for the areas of the coiled tubing surface. We exploit a Hierarchical Allocation Strategy (HAS) to preserve sample stratification when generating any desirable sample numbers. This permits us to reduce variances when applying our algorithm to Monte Carlo Direct Lighting for realistic image generation. We accelerate the sampling process using a segmentation technique in the integration domain. Our algorithm thus runs 324 orders of magnitude faster when using faster SSA algorithm where the order of the magnitude is proportional to the sample numbers. Finally, we employ a parabolic interpolation technique to decrease the average errors occurred for using the segmentation technique. This permits us to produce nearly constant average errors, independent of the sample numbers. The proposed algorithm is novel, efficient in computing and feasible for realistic image generation using Monte Carlo method.
ER -