基于相对梯度正则化的Retinex变分模型及其应用

doi:10.11959/j.issn.1000-436x.2017212

摘要/Abstract

摘要：

针对全变分 Retinex 模型使用反射分量的全变分作为正则项存在的不足，引入相对梯度并构建扩展的全变分正则项，提出一种新的Retinex变分模型。相比于变分Retinex模型、全变分Retinex模型，该模型获取的照度分量更加光滑，同时反射分量能够分辨更多结构信息和细节要素。进而，提出一种综合考虑照度分量和反射分量的图像增强模型。通过调整模型参数，可有效应用于高动态范围图像色调映射、非均匀照度增强等图像处理领域。与其他算法的对比显示，该图像增强模型能够有效处理上述问题并取得较好的效果。

关键词: Retinex变分模型, 相对梯度, 综合增强模型, 高动态范围图像色调映射

Abstract:

In view of the shortcomings of the total variational Retinex model which use the total variation (TV) of the reflection as the regularization.An extension of TV regularization with the concept of relative gradient was introduced and finally a new variational Retinex model was proposed.Compared with variational Retinex and total variational Retinex model,the proposed model can preserve the estimated reflectance with more details as well as the more smoothed illumination.Further,a new integrated image enhancement model considering both the illumination and the reflectance was proposed.By adjusting the model parameters,the proposed model can be effectively applied to high dynamic range image tone mapping and non-uniform illumination enhancement.Compared with other algorithms,the proposed model can better handle the above image enhancement problems.

Key words: variational Retinex model, relative gradient, integrated image enhancement model, HDR

中图分类号:

TP391.4

智宁,毛善君,李梅. 基于相对梯度正则化的Retinex变分模型及其应用[J]. 通信学报, 2017, 38(11): 65-75.

Ning ZHI,Shan-jun MAO,Mei LI. Variational Retinex model based on an extension of TV regularization with relative gradient and its application[J]. Journal on Communications, 2017, 38(11): 65-75.

图/表 10

图1

图2

图3

图4

图5

图6

表1

表2

图7

表3

参考文献 31

[1]	LAND E H , MCCANN J J . Lightness and Retinex theory[J]. Journal of the Optical Society of America, 1971,61(61): 1-11.
[2]	LAND E H . An alternative technique for the computation of the designator in the Retinex theory of color vision[J]. The National Academy of Sciences, 1986,83(10): 3078-3080.
[3]	BRAINARD D H , WANDELL B A . Analysis of the Retinex theory of color vision[J]. Journal of the Optical Society of America A Optics ＆Image Science, 1986,3(10):1651.
[4]	RIZZI A . Designator Retinex,milano Retinex and the locality issue[J]. Electronic Imaging, 2016,2016(6): 1-5.
[5]	LAND E H . Recent advances in Retinex theory and some implications for cortical com-putations:color vision and the natural image[J]. Proceedings of the National Academy of Sciences of the United States of America, 1983,80(16): 5163-5169.
[6]	PROVENZI E , FIERRO M , RIZZI A ,et al. Random spray Retinex:a new Retinex implementation to investigate the local properties of the model[J]. IEEE Transactions on Image Processing, 2007,16(1): 162-171.
[7]	KOL?S ? , FARUP I , RIZZI A . Spatio-temporal Retinex-inspired envelope with stochastic sampling:a framework for spatial color algorithms[J]. Journal of Imaging Science ＆ Technology, 2011,55(4): 40503-1-40503-10.
[8]	JOBSON D J , RAHMAN Z U , WOODELL G A . A multiscale Retinex for bridging the gap between color images and the human observation of scenes[J]. IEEE Transactions on Image Processing, 1997,6(7): 965-976.
[9]	JOBSON D J , RAHMAN Z , WOODELL G A . Properties and performance of a center/surround Retinex[J]. IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society, 1997,6(3): 451-62.
[10]	MOREL J M,PETRO , SBERT C . A PDE formalization of Retinex theory[J]. IEEE Transactions on Image Processing a Publication of the IEEE Signal Processing Society, 2010,19(11): 2825-2837.
[11]	KIMMEL R , ELAD M , SHAKED D ,et al. A variational framework for Retinex[J]. International Journal of Computer Vision, 2003,52(1): 7-23.
[12]	MA W . Variational models in image and signal enhancement[J]. Dissertation ＆ Theses-Gradworks, 2011
[13]	MA W Y , OSHER S ,et al. A TV bregman iterative model of retinex theory[J]. Inverse Problems ＆ Imaging, 2012,4(4): 697-708.
[14]	MA W , MOREL J M , OSHER S ,et al. An L1-based variational model for Retinex theory and its application to medical images[C]//CVPR 2011:153-160. CVPR, 2011: 153-160.
[15]	NG M K , WANG W . A total variation model for Retinex[J]. Siam Journal on Imaging Sciences, 2011,4(1): 345-365.
[16]	WANG W K . NG M . A nonlocal total variation model for image decomposition:illumination and reflectance[J]. Numerical Mathematics Theory Methods ＆ Applications, 2014,7(3): 334-355.
[17]	ZOSSO D , TRAN G , OSHER S J . Non-local Retinex-a unifying framework and beyond[J]. SIAM Journal on Imaging Sciences, 2015,8(2): 787-826.
[18]	DOU Z Y , GAO K , ZHANG B ,et al. Realistic image rendition using a variable exponent functional model for Retinex[J]. Sensors, 2016,16: 1-16.
[19]	LIU H , XIONG R , ZHANG X ,et al. Non-local gradient sparsity regularization for image restoration[J]. IEEE Transactions on Circuits ＆Systems for Video Technology, 2016:1.
[20]	WANG Y , LIU H , WU Z . Towards relative gradient and its applications[C]// Sixth International Conference on Graphic and Image Processing (ICGIP 2014). International Society for Optics and Photonics, 2015.
[21]	GOLDSTEIN T , OSHER S . The split bregman method for l1-regularized problems[J]. SIAM Journal on Imaging Sciences, 2009,2(2): 323-343.
[22]	GUO X , YU L , LING H . LIME:Low-light image enhancement via illumination map estimation[J]. IEEE Transactions on Image Processing a Publication of the IEEE Signal Processing Society, 2016,26(2): 982-993.
[23]	SHIN Y , JEONG S , LEE S . Efficient naturalness restoration for non-uniform illumination images[J]. IET Image Processing, 2015,9(8): 662-671.
[24]	WANG S , ZHENG J , HU H M ,et al. Naturalness preserved enhancement algorithm for non-uniform illumination images[J]. IEEE Transactions on Image Processing a Publication of the IEEE Signal Processing Society, 2013,22(9): 3538-3548.
[25]	XIAOMENG Y U , CHANG H , QUANBAN L I ,et al. Improved multi-scale Retinex image enhancement algorithm[J]. Journal of Computer Applications, 2010,30(8): 2091-2093.
[26]	刘志成, 王殿伟, 刘颖 ,等. 基于二维伽马函数的光照不均匀图像自适应校正算法[J]. 北京理工大学学报, 2016,36(2): 191-196.
	LIU Z C , WANG D W , LIU Y ,et al. Adaptive adjustment algorithm for non-uniform illumination images based on 2D gamma function[J]. Transactions of Beijing Institute of Technology, 2016(2): 191-196.
[27]	FARBMAN Z , FATTAL R , LISCHINSKI D ,et al. Edge-preserving decompositions for multi-scale tone and detail manipulation[J]. ACM Transactions on Graphics, 2008,27(3): 15-19.
[28]	FATTAL R , LISCHINSKI D , WERMAN M . Gradient domain high dynamic range compression[J]. ACM Transactions on Graphics, 2002,21(3): 249-256.
[29]	PARIS S , HASINOFF S W , KAUTZ J . Local Laplacian filters:edge-aware image processing with a Laplacian pyramid[J]. ACM Transactions on Graphics, 2011,30(4): 1244-1259.
[30]	TUMBLIN J , . Low curvature image simplifiers (LCIS):a boundary hierarchy for detail-preserving contrast reduction[C]// Conference on Computer Graphics ＆ Integrative Techniques. 1999: 83-90.
[31]	都琳, 孙华燕, 王帅 ,等. 针对动态目标的高动态范围图像融合算法研究[J]. 光学学报, 2017(4): 109-117.
	DU L , SUN H Y , WANG S ,et al. High dynamic range image fusion algorithm for moving targets[J]. Acta Optica Sinica, 2017(4): 109-117.

图序	entropy	ALSD	AG
图5(a)	7.13	22.62	3.97
图5(b)	7.59	34.51	7.02
图5(c)	7.42	34.14	4.39
图5(d)	7.55	30.70	5.05
图5(e)	7.45	34.85	5.74
图5(f)	7.70	42.58	7.08
图6(a)	6.84	18.95	7.82
图6(b)	7.51	37.71	17.98
图6(c)	6.99	25.14	8.42
图6(d)	7.70	32.97	14.00
图6(e)	7.28	31.25	13.86
图6(f)	7.71	41.62	17.34

图序	LIME	MSRCR	ENR	NPE	本文算法
图5	4.360 0	13.370 0	1.903 2	226.140 0	12.500 0
图6	0.826 8	3.260 4	0.577 2	38.750 0	2.010 0

算法	entropy	ALSD	平均亮度	处理时间/s
原图像	6.71	30.12	73.4	—
Fattal	7.19	47.83	119.59	87.1
LLF	7.1	36.34	85.74	245.6
WLS	6.72	34.54	59.85	27.2
本文算法	7.58	55.73	117.96	10.1