用sure引导自适应采样与采样颜色直方图重构-计算机辅助设计与.pdf

第28 卷 第4 期 计算机辅助设计与图形学学报 Vol. 28 No.4 2016 年4 月 Journal of Computer-Aided Design Computer Graphics Apr. 2016 用SURE 引导自适应采样与采样颜色直方图重构 1,2) 1,2) 1) 袁红亮 , 吴付坤 , 郑昌文 1) ( 中国科学院软件研究所天基综合信息系统重点实验室 北京 100190) 2) ( 中国科学院大学 北京 100190) ) : 针对现有采样颜色直方图重构缺少自适应采样方法的问题, 提出一种基于 Stein’s unbiased risk estima- tor(SURE) 的像素均方误差计算方法. 首先对图像空间粗采样; 然后在GPU 上分块计算SURE 来评估图像每个像素的 误差度量值, 并用该值引导非均匀自适应采样; 最后使用非局部多尺度滤波方法重构图像. 实验结果表明, 不论从 图像噪声定量化指标还是视觉效果来看, 该方法都极大地改进了图像绘制质量. ：GPU 加速; 自适应采样; SURE; 均方误差 ：TP391.41 Adaptive Sampling Guided by SURE and Sample Color Histogram Reconstruction Yuan Hongliang1,2), Wu Fukun1,2), and Zheng Changwen1) 1) (Science and Technology on Integrated Information System Laboratory, Institute of Software, Chinese Academy of Sciences, Beijing 100190) 2) ( University of Chinese Academy of Sciences, Beijing 100190) Abstract: We present an approach for computing pixel-wise mean squared error based on Stein’s unbiased risk estimator (SURE), to perform adaptive sampling for histogram-based sample reconstruction in Monte Carlo path tracing. Firstly, the image space is coarsely sampled; then we compute SURE on GPU by block estimates error metric of each pixel, to guide subsequent non-uniform adaptive sampling. Finally, nonlocal multi-scale filtering is used to reconstruct image. The experimental results demonstrate substantial improvements in terms of both nu- merical error and visual quality. Key words: GPU acceleration; adaptive sampling; Stein’s unbiased risk estimator; mean squared error , Monte Carlo(MC) 2 MC . . . , , . MC ,