蒙卡渲染画面多特征非局部均值滤波降噪算法

doi:10.16182/j.issn1004731x.joss.20-1037

系统仿真学报 ›› 2022, Vol. 34 ›› Issue (6): 1259-1266.doi: 10.16182/j.issn1004731x.joss.20-1037

蒙卡渲染画面多特征非局部均值滤波降噪算法

杨凯(), 陈纯毅(), 胡小娟, 于海洋

长春理工大学计算机科学技术学院，吉林长春 130022

收稿日期:2020-12-23 修回日期:2021-01-28 出版日期:2022-06-30 发布日期:2022-06-16
通讯作者: 陈纯毅 E-mail:741910453@qq.com;chenchunyi@hotmail.com
作者简介:杨凯(1997-)，男，硕士生，研究方向为计算机图形学。E-mail：741910453@qq.com
基金资助:
国家自然科学基金(U19A2063);吉林省科技发展计划(20190302113GX);吉林省教育厅“十三五”科学技术研究项目(JJKH20200792KJ)

Denoising Algorithm Based on Multi-feature Non-local Mean Filtering for Monte Carlo Rendered Images

Kai Yang(), Chunyi Chen(), Xiaojuan Hu, Haiyang Yu

School of Computer Science and Technology, Changchun University of Science and Technology, Changchun 130022, China

Received:2020-12-23 Revised:2021-01-28 Online:2022-06-30 Published:2022-06-16
Contact: Chunyi Chen E-mail:741910453@qq.com;chenchunyi@hotmail.com

摘要/Abstract

摘要：

针对蒙特卡罗渲染在低光线路径采样率下绘制的图像容易出现噪点的问题，提出一种蒙特卡罗渲染画面多特征非局部均值降噪算法。使用canny算法对场景反射率图进行梯度化，以反射率梯度图为引导图，利用引导滤波器对法向量图预滤波；求出预滤波后法向量图中图像块之间的结构相似性，利用结构相似性倒数的对数值对非局部均值滤波器权值进行改进；利用改进后的非局部均值滤波器对噪点图像进行滤波重构。实验结果表明：该算法能在典型场景中有效降低渲染画面的噪声，改善均方误差和峰值信噪比质量指标。

关键词: 蒙特卡罗, 渲染, 引导滤波器, 非局部均值滤波器, 结构相似性

Abstract:

Aiming at the rendering noise in Monte Carlo synthesized images induced by the low light-path sampling rate, a denoising algorithm based on the multi-feature non-local-mean filtering is proposed. The gradient image of the scene's albedo information is calculatedwith the canny operator, and a guided filter together with the said gradient image is employed to prefilter the normal vector image. The structural similarity of the sub-blocks in the prefiltered normal vector image is calculated and the improved weights of the non-local mean filter are computed according to the logarithmic value of the reciprocal of the structural similarity. The improved non-local mean filter is used to implement the reconstruction of the noisy image. The experimental results show that the proposed algorithm can effectively reduce the level of Monte Carlo rendering noise in typical scenes, and can improve the metrics of mean square error and peak signal-to-noise ratio.

Key words: Monte Carlo, rendering, guide filter, non-local mean filter, structural similarity

中图分类号:

TP391.9

杨凯, 陈纯毅, 胡小娟, 于海洋. 蒙卡渲染画面多特征非局部均值滤波降噪算法[J]. 系统仿真学报, 2022, 34(6): 1259-1266.

Kai Yang, Chunyi Chen, Xiaojuan Hu, Haiyang Yu. Denoising Algorithm Based on Multi-feature Non-local Mean Filtering for Monte Carlo Rendered Images[J]. Journal of System Simulation, 2022, 34(6): 1259-1266.

图/表 10

图1

图2

图3

表1

图4

图5

图6

表2

表3

表4

参考文献 21

1	Overbeck R S, Donner C, Ramamoorthi R. Adaptive Wavelet Rendering[J]. ACM Transactions on Graphics (S0730-0301), 2009, 28(5): 1-12.
2	Rousselle F, Knaus C, Zwicker M. Adaptive Sampling and Reconstruction Using Greedy Error Minimization[J]. ACM Transactions on Graphics (S0730-0301), 2011, 30(6): 1-12.
3	Rousselle F, Knaus C, Zwicker M. Adaptive Rendering with Non-Local Means Filtering[J]. ACM Transactions on Graphics(S0730-0301), 2012, 31(6): 1-11.
4	Delbracio M, Muse P, Buades A, et al. Boosting Monte Carlo Rendering by Ray Histogram Fusion[J]. ACM Transactions on Graphics(S0730-0301), 2014, 33(1): 1-15.
5	Lehtinen J, Aila T, Chen J W, et al. Temporal Light Field Reconstruction for Rendering Distributed Effects[J].ACM Transactions on Graphics(S0730-0301), 2011, 30(4): 1-12.
6	Mara M, Mcguire M, Bitterli B, et al. An Efficient Denoising Algorithm for Global Illumination[C]// High Performance Graphics. ACM: 2017: 1-7.
7	Belcour Laurent, Bala Kavita, Soler Cyril. A Local Frequency Analysis of Light Scattering and Absorption[J]. ACM Transactions on Graphics(S0730-0301), 2014, 33(5): 1-17.
8	Moon B, Iglesias-Glesias-Guitian J A, McDonagh S. Noise Reduction on G-Buffers for Monte Carlo fiLtering[J]. Computer Graphics Forum (S0167-7055), 2017, 36(8): 600-612.
9	Bitterli B, Rousselle F, Moon B. Nonlinearly Weighted Firstorder Regression for Denoising Monte Carlo Renderings[J]. Comput Graph Forum(S1467-8659), 2016, 35(4): 107-117.
10	Boughida M, Boubekeur T. Bayesian Collaborative Denoising for Monte Carlo Rendering[J]. Comput Graphics Forum(S1467-8659), 2017, 36(4): 137-153.
11	Yang Xin, Wang Dawei, Hu Wenbo, et al. Fast Reconstruction for Monte Carlo Rendering Using Deep Convolutional Networks[J]. IEEE Access(S1531-1309), 2019, 7: 21177-21187.
12	Kettunen M, Hrknen E, Lehtinen J. Deep Convolutional Reconstruction for Gradient-Domain Rendering[J]. ACM Transactions on Graphics (S0730-0301), 2019, 38(4): 1-12.
13	Gharbi M, Li T M, Aittala M, et al. Sample-Based Monte Carlo Denoising Using a Kernel-Splatting Network[J]. ACM Transactions on Graphics(S0730-0301), 2019, 38(4): 1-12.
14	Bauszat Pablo, Eisemann Martin, Eisemann Elmar. General and Robust Error Estimation and Reconstruction for Monte Carlo Rendering[J]. Computer Graphics Forum(S0167-7055), 2015, 34(2): 597-608.
15	Moon B, McDonagh S, Mitchell K, et al. Adaptive Polynomial Rendering[J]. ACM Transactions on Graphics(S0730-0301), 2016, 35(4): No. 40.
16	Vicini Delio, Adler David, Novák Jan, et al. Denoising Deep Monte Carlo Renderings[J]. Computer Graphics Forum (S0167-7055), 2018, 38(4): 316-327.
17	Yuan Yaoshen, Yu Leiming, Fang Qianqian. Denoising in Monte Carlo Photon Transport Simulation Using GPU-Accelerated Adaptive Non-Local Mean Filter[C] // Clinical and Translational Biophotonics. 2018.
18	Mazi M, Vicini D, Zwicker M. Regularizing Imager Econstruction for Gradient-Domain Rendering with Feature Patches[J]. Computer Graphics Forum (S0167-7055), 2016, 35(2): 263-273.
19	He K, Sun J, Tang X. Guided Image Filtering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S1939-3539), 2010, 35(6): 1397-1409.
20	Buades A, Coll B, Morel J. A Review of Image Denoising Algorithms, with a New One [J]. SIAM Journal on Multiscale Modeling and Simulation (S0036-1445), 2005, 4(2): 490-530.
21	Wang Z, Bovik A C, Sheikh H R, et al. Image Quality Assessment: From Error Visibility to Structural Similarity[J]. IEEE Transaction on Image Processing (S1057-7149), 2004, 13(4): 600-612.

搜索框与相似块	PSNR	MSE
5×5、3×3	33.584 6	0.001 20
7×7、5×5	35.226 7	0.000 98
11×11、5×5	33.660 3	0.001 20
21×21、7×7	34.055 8	0.001 30

算法	PSNR	MSE
MC	28.358 7	0.004 30
NLM	30.066 2	0.003 20
文献[11]	34.590 7	0.000 34
本文	36.009 3	0.000 74

算法	PSNR	MSE
MC	27.482 9	0.005 50
NLM	32.742 1	0.003 50
文献[11]	34.413 0	0.000 86
本文	34.495 2	0.001 10

算法	PSNR	MSE
MC	24.843 7	0.009 90
NLM	30.066 2	0.002 90
文献[11]	32.894 5	0.000 51
本文	34.895 4	0.000 98

[1]	刘逊韵, 徐新海, 朱成璋, 李豪, 曾蕾. 面向多视角渲染任务的超实时可视仿真系统[J]. 系统仿真学报, 2022, 34(7): 1619-1628.
[2]	程敬海, 汪红志, 徐罗元, 夏天. 基于蒙特卡罗法的放射测量系统建模与仿真实现[J]. 系统仿真学报, 2021, 33(11): 2663-2672.
[3]	刘灏哲, 李伟, 马萍, 杨明. 基于多源先验数据的系统性能评估方法[J]. 系统仿真学报, 2021, 33(11): 2673-2680.
[4]	孙炳利, 宋晓, 龚光红. 基于供应商画像与马尔可夫蒙特卡罗仿真的供应商选择[J]. 系统仿真学报, 2021, 33(11): 2720-2732.
[5]	刘佳哲, 陈纯毅, 胡小娟, 梁伟东, 邢琦玮, 杨华民. 面向手机的云端VR三维场景立体显示与交互框架[J]. 系统仿真学报, 2020, 32(7): 1360-1374.
[6]	邵威, 刘畅, 贾金原. 基于光照贴图的Web3D全局光照协作式云渲染系统[J]. 系统仿真学报, 2020, 32(4): 649-659.
[7]	刘畅, 贾金原, 陆一凡, 张乾, 赵磊. 光图树：Web3D动态场景全局光照的协同实时渲染[J]. 系统仿真学报, 2019, 31(8): 1591-1604.
[8]	刘海洋, 胡晓峰, 雷旭. 基于图形集群的远程实时渲染系统研究[J]. 系统仿真学报, 2019, 31(5): 886-892.
[9]	李娟, 张昆玉, 李海波, 杨莉娟, 王蒙迪. 基于蒙特卡罗法的AUV任务可靠性分析[J]. 系统仿真学报, 2019, 31(10): 2131-2137.
[10]	高山晓, 吴献. 一种基于GPU构造阴影线的软阴影绘制算法[J]. 系统仿真学报, 2018, 30(4): 1279-1286.
[11]	周守东. 基于Woodcock Tracking的快速蒙特卡罗直接体绘制方法[J]. 系统仿真学报, 2017, 29(5): 1125-1131.
[12]	翟宇梅, 赵瑞星. 耿贝尔分布的最优线性矩估计[J]. 系统仿真学报, 2016, 28(5): 1070-1076.
[13]	谢永华, 袁复兴, 王畅. 基于重要性采样的三维云光照模型的研究[J]. 系统仿真学报, 2016, 28(1): 57-62.
[14]	王圣华, 谭剑. 中国皮影体可视化仿真方法研究[J]. 系统仿真学报, 2015, 27(9): 2126-2134.
[15]	万明, 梁霞, 张凤鸣. 一种四叉树地形渲染裂缝的改进消除算法[J]. 系统仿真学报, 2015, 27(7): 1520-1525.

蒙卡渲染画面多特征非局部均值滤波降噪算法

Denoising Algorithm Based on Multi-feature Non-local Mean Filtering for Monte Carlo Rendered Images

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 21

相关文章 15

编辑推荐

Metrics

本文评价