Improved Algorithm for Removing Highlights with Guided Filtering

doi:10.3778/j.issn.1002-8331.2006-0415

Abstract

Abstract:

In order to solve the problems of discoloration of black pixels, loss of edges and textures, and artifacts when removing highlights, an improved algorithm for removing highlights with guided filtering is proposed. Firstly, it separates the black pixels and other pixels by setting the first threshold, and estimates the maximum diffuse chromaticity of the black pixels and other pixels separately. These can avoid the black pixels from fading. Secondly, it smoothes the maximum chromaticity image by using the guided filter. The artifact effect is avoided. Finally, the pixels are divided into two categories according to the relationship between the denominator and the second threshold. Moreover the two categories are used different methods to separate diffuse reflection components. It avoids error separation caused by denominator 0. Experimental results show that the algorithm is superior to traditional algorithms in processing black pixels, preserving edges and textures, and avoiding artifacts.

Key words: highlight removal, guided filter, double-color?reflection, specular reflection, diffuse reflection

摘要：

为了解决传统的高光去除方法在去除图像高光时黑色像素点褪色、边缘和纹理缺失、产生伪影效应的问题，提出了一种导向滤波的高光去除改进算法。该算法通过设置第一阈值分离出图像中的黑色像素和其他像素，并采用不同方法分别估算黑色像素和其他像素的最大漫反射色度，避免黑色像素点发生褪色；利用导向滤波器对最大色度图进行了平滑处理，避免了伪影效应；在分离漫反射分量过程中，根据分母与第二阈值的关系，将滤波后图像中的像素点划分为两类，并对不同类别的像素点采用不同方法实现漫反射分量的分离，避免了分母为0导致错误分离。根据实验结果，该算法相比于传统算法，在防止黑色像素褪色、保留图像边缘和纹理信息、避免伪影方面均表现出优势。

关键词: 高光去除, 导向滤波, 双色反射模型, 镜面反射, 漫反射

GUO Biru, KONG Weiwei, CHEN Bin. Improved Algorithm for Removing Highlights with Guided Filtering[J]. Computer Engineering and Applications, 2021, 57(20): 229-235.

郭碧茹，孔韦韦，陈斌. 导向滤波的高光去除改进算法[J]. 计算机工程与应用, 2021, 57(20): 229-235.

References

[1] 蔡清源.三维物体表面的反射特性恢复[D].杭州：浙江大学，2010.
CAI Qingyuan.Recovering reflections of three-dimesional object surfaces[D].Hangzhou：Zhejiang University，2010.
[2] LI C，LIN S，ZHOU K.Specular highlight removal in facial images[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2017：2780-2789.
[3] 芦彦欣.面向金属零件特征提取的高光去除技术研究与实现[D].杭州：浙江大学，2019.
LU Yanxin.Research and implementation of highlight removal technology for feature extraction of metal parts[D].Hangzhou：Zhejiang University，2019.
[4] FERIS R，RASKAR R，TAN K H，et al.Specular reflection reduction with multi-flash imaging[C]//17th Brazilian Symposium on Computer Graphics and Image Processing，2004：316-321.
[5] SHAFER S.Using color to separate reflection components[M]//Color.[S.l.]：Jones and Bartlett Publishers，Inc，1992.
[6] 卢桂荣，汤景凡，姜明.基于快速双边滤波的图像高光去除研究[J].计算机工程与应用，2014，50（10）：176-179.
LU Guirong，TANG Jingfan，JIANG Ming.Research on image highlight removal based on fast bilateral filter[J].Computer Engineering and Applications，2014，50（10）：176-179.
[7] TAN R T，IKEUCHI K.Separating reflection components of textured surfaces using a single image[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2005，27（2）：178-193.
[8] 郑芝寰，沈会良，程岳.基于单幅图像的高光分量可靠去除方法[J].杭州电子科技大学学报，2011，31（4）：165-168.
ZHENG Zhihuan，SHEN Huiliang，CHENG Yue.Reliable specular highlight removal in a single image[J].Journal of Hangzhou Dianzi University，2011，31（4）：165-168.
[9] YANG Qingxiong，WANG Shengnan，AHUJA N.Real-time specular highlight removal using bilateral filtering[C]//European Conference on Computer Vision（ECCV），2010：87-100.
[10] 高如新，李雪颖.双边滤波的改进高光去除[J].中国图象图形学报，2018，23（1）：9-17.
GAO Ruxin，LI Xueying.Specular highlight removal using improved bilateral filtering[J].Journal of Image and Graphics，2018，23（1）：9-17.
[11] GAVASKAR R G，CHAUDHURY K N.Fast adaptive bilateral filtering[J].IEEE Transactions on Image Processing，2018，28（2）：779-790.
[12] 许丽，宋滢.高光去除的聚类算法改进[J].计算机系统应用，2020，29（1）：209-214.
XU Li，SONG Ying.Specular highlight removal using improved pixel clustering[J].Computer Systems & Applications，2020，29（1）：209-214.
[13] SHEN Huiliang，ZHENG Zhihuan.Real-time highlight removal using intensity ratio[J].Applied Optics，2013，52（19）：4483-4493.
[14] HE Kaiming，SUN Jian.Image completion approaches using the statistics of similar patches[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2014，36（12）：2423-2435.
[15] REN Weihong，TIAN Jiandong，TANG Yandong.Specular reflection separation with color-lines constraint[J].IEEE Transactions on Image Processing，2017，26（5）：2327-2337.