基于变分模型的梯度域色阶映射算法

doi:10.11959/j.issn.1000-436x.2015001

摘要/Abstract

摘要：

针对梯度域动态范围压缩算法存在伪边缘以及局部细节扭曲等现象，提出了一种基于变分模型的梯度域色阶映射方法。首先，在梯度域内构造了一个既能够压缩图像的动态范围又能够保证边缘与细节信息的变分模型；其次，将Gibbs采样的思想引入到最速下降法中，在求解变分模型最优解的同时有效地提高了无约束最优算法的收敛速度；最后通过改进的最速下降法得到变分模型最优解。实验结果表明该算法能够有效地去除光晕，得到细节保持完好的低动态范围图像。另外，改进的最速下降法保证了算法的实时性。

关键词: 高动态范围图像, 梯度域, 变分模型, 最速下降法, Gibbs采样, 泊松方程

Abstract:

Due to the low dynamic range image which was generated by the gradient domain high dynamic range compression algorithm contain the artificial boundaries and local detail distortions,a variational model in gradient domain was proposed to improve the performance of the traditional algorithm.First of all,a variational model in gradient domain was introduced to compress dynamic range,meanwhile details and edges were preserved simultaneously.Afterwards,the rate of convergence was improved by introducing the ideology of Gibbs sampler.Eventually,the improved method was employed to obtain the optimal solution of the variational model.Experimental results demonstrate that proposed algorithm reduces the degree of artificial boundaries,meanwhile low dynamic range image represents excellent capacity of detail preservation.Moreover,the real-time performance is guaranteed by the improved steepest descent method.

Key words: high dynamic range image, gradient domain, variational model, steepest descent method, Gibbs sampler, Poisson equation

席志红,赵蓝飞,张驰,张志民. 基于变分模型的梯度域色阶映射算法[J]. 通信学报, 2015, 36(1): 1-8.

IZhi-hong X,Lan-fei ZHAO,Chi ZHANG,Zhong-min ZHANG. Tone mapping based on variational model in gradient domain[J]. Journal on Communications, 2015, 36(1): 1-8.

图/表 9

图1

图2

图3

图4

图5

图6

表1

图7

表2

参考文献 19

[1]	杨克虎，姬靖，郭建军等．高动态范围图像和色阶映射算子[J]．自动化学报， 2009,35(2):114-122. YANG K H , JI J , GUO J J , et al. High dynamic range images and tone mapping operator[J]. Acta Automatica Sinica, 2009,35(2):114-122.
[2]	王作省，邹少芳，王章野．高动态图像色调映射技术新进展[J]．计算机应用研究， 2010,27(7):2421-2424. WANG Z S , ZOU S F , WANG Z Y . Recent tone mapping techniques for high dynamic range images[J]. Application Research of Computers, 2010,27(7):2421-2424.
[3]	赵蓝飞，席志红．基于Mean Shift 的光晕抑制算法[J]．系统工程与电子技术， 2013,35(7):1558-1563. ZHAO L F , XI Z H . Halo inhibition algorithm based on mean shift filter[J]. Journal of Systems Engineering and Electronics, 2013,35(7):1558-1563.
[4]	FATTAL R , LISEHINSKI D , WERMAN M . Gradient domain high dynamic range compression[J]. ACM Transactions on Graphics, 2002,21(3): 249-256.
[5]	AUBERT G , VESE L . A variational method in image recovery[J]. SIAM Journal of Numerical Analysis, 1997,34(5): 1948-1979.
[6]	BERTSEKAS D . Non-Linear Programming[M]. Belmont: Athena Scientific, 1999.
[7]	BOYD P , VANDENBERGHE L . Convex Optimization[M]. New York: Cambridge University Press, 2004.
[8]	DAI Y H , YUAN Y . A nonlinear conjugate gradient method with a strong global convergence property[J]. SIAM Journal of Optimization, 1999,10(1): 177-182.
[9]	袁亚湘，孙文瑜．最优化理论与方法[M]．北京：科学出版社， 2001. YUAN Y X , SUN W Y . Theory and Methods of Optimization[M]. Beijing: Science Press, 2001.
[10]	KOZUMI H , KOBAYASHI G . Gibbs sampling methods for Bayesian quantile regression[J]. Journal of Statistical Computation and Simulation, 2011,81(11): 1565-1578.
[11]	MEERSCHAERT M M , TADJERAN C . Finite difference approximations for fractional advection-dispersion flow equations[J]. Journal Computational and Applied Mathematics, 2004,172(1): 65-77.
[12]	席志红，赵蓝飞．基于格林函数的动态范围压缩算法[J]．计算机辅助设计与图形学学报， 2014,26(4):581-589. XI Z H , ZHAO L F . Dynamic range compression algorithm based on Green's function[J]. Journal of Computer-Aided Design ＆; Computer Graphics, 2014,26(4):581-589.
[13]	JOBSON D J , RAHMAN Z , WOODELL G A . Properties and performance of the center/surround retinex[J]. IEEE Transactions on Image Processing, 1997,6(3): 451-462.
[14]	DURAND F , DORSEY J . Fast bilateral filtering for the display of high dynamic range images[J]. ACM Transactions on Graphics, 2002,21(3): 257-266.
[15]	DRAGO F , MYSZKOWSKI K , ANNEN T , et al. Adaptive logarithmic mapping for displaying high contrast scenes[J]. Computer Graphics Forum, 2003,22(3): 419-426.
[16]	REINHARD E , DEVLIN K . Dynamic range reduction inspired by photoreceptor physiology[J]. IEEE Transactions on Visualization and Computer Graphics, 2005,11(1): 13-24.
[17]	WANG Z , BOVIK A C , SHEIKH H R . Image quality assessment:from error visibility to structural similarity[J]. IEEE Transactions on Image Processing, 2004,13(4): 600-612.
[18]	范媛媛，沈湘衡，桑英军．基于对比度敏感度的无参考图像清晰度评价[J]．光学精密工程， 2011,19(10):2485-2493. FAN Y Y , SHEN X H , SANG Y J . No reference image sharpness assessment based on contrast sensitivity[J]. Optics and Precision Engineering, 2011,19(10):2485-2493.
[19]	SHEIKH H R , BOVIK A C . Image information and visual quality[J]. IEEE Transactions on Image Processing, 2006,15(2): 430-444.

图像名称	改进前	改进后
church	23	11
sunset	18	9
rosette	24	11
studio	15	8

算法名称	MSSIM	CFSNRS	VIF
直接显示	0.953	0.425	0.387
MSR	0.834	0.792	0.568
Bilateral	0.771	0.645	0.544
Photoreceptor	0.795	0.501	0.625
Drago	0.801	0.587	0.687
梯度衰减	0.878	0.784	0.656
本文算法	0.899	0.802	0.865