没有合适的资源?快使用搜索试试~ 我知道了~
Colorization for gray scale facial image by locality-constrained...
0 下载量 179 浏览量
2021-02-09
00:00:29
上传
评论
收藏 669KB PDF 举报
温馨提示
Colorization for gray scale facial image is an important technique in various practical applications. However, the methods that have been proposed are essentially semi-automatic. In this paper, we present a new probabilistic framework based on Maximum A Posteriori (MAP) estimation to automatically transform the given gray scale facial image to corresponding color one. Firstly, the input image is divided into several patches and non-parametric Markov random field (MRF) is employed to formulate th
资源推荐
资源详情
资源评论
Colorization for Gray Scale Facial Image
by Locality-Constrained Linear Coding
Yang Liang, Mingli Song, Jiajun Bu, and Chun Chen
Zhejiang Provincial Key Laboratory of Service Robot
College of Computer Science, Zhejiang University
{liangyang,brooksong,bjj,chenc}@zju.edu.cn
Abstract. Colorization for gray scale facial image is an important tech-
nique in various practical applications. However, the methods that have
been proposed are essentially semi-automatic. In this paper, we present a
new probabilistic framework based on Maximum A Posteriori (MAP) es-
timation to automatically transform the given gray scale facial image to
corresponding color one. Firstly, the input image is divided into several
patches and non-parametric Markov random field (MRF) is employed to
formulate the global energy. Secondly, Locality-constrained Linear Cod-
ing (LLC) is employed to learn the color distribution for each patch.
At the same time, the simulated annealing algorithm is employed to
iteratively update the patches chosen by LLC to optimize the MRF by
decreasing global energy cost. The experimental results demonstrate that
the proposed framework is effective to colorize the gray scale facial im-
ages to corresponding color ones.
Keywords: Colorization,MAP,MRF,LLC.
1 Introduction
Human face conveys lots of information such as identity, appearance, etc. during
social communication or in security systems. However, sometimes only gray scale
images or portraits can be obtained whose facial details like skin color, lip luster,
etc. are lost. For example, because of the quality of surveillance cameras or
storage space restriction, only gray scale images may be obtained. The same for
archeologists, they may only obtain historical gray scale images. So colorization
for gray scale facial image is useful.
In 1970 Wilson Markle[1] introduced the term colorization to describe the
computer assisted process of adding color to a gray scale image. Several ap-
proaches have been presented towards successful image colorization. In general,
the existing colorization approaches can be classified into two groups, one is
example-based method and the other is optimization-based method.
The example-based method colorizes the gray scale image by choosing a sim-
ilar image as a reference, and transferring its colors to the input gray scale
image[2][3]. Welsh in [2] transfers the entire color mood by matching the lumi-
nance and texture of the example image to the target gray scale image pixel by
W. Lin et al. (Eds.): PCM 2012, LNCS 7674, pp. 57–67, 2012.
c
Springer-Verlag Berlin Heide lberg 2012
58 Y. Liang et al.
pixel. But it always fails on those images that are hardly segmented by luminance
or texture. Kekre and Thepade in [3] searches the reference color image for the
same palette of a certain scale pixel window by matching the luminance values
of reference color image to target gray scale image. However, a suitable reference
color image whose color mood and composition are required to be similar to the
target gray scale image may take effort to find and it may fail on those images
that are hardly segmented by luminance or texture.
The optimization-based method colorizes the image based on the color label
priors offered by users [4][5][6]. Levin in [4] proposes a premise that neighboring
pixels will have similar colors if they have similar intensities and Levin formalizes
the premise as an optimization problem using a quadratic cost function based
on a few color labels figured by users. The method can produce comparatively
better colorized image than the techniques of former group without precise image
segmentation, but it still needs user interactions to confirm several color labels
for different color regions and dramatically depends on the accuracy of the color
labels as prior.
Obviously, all those mentioned colorization techniques require human interac-
tions to confirm either the color mood from a reference image or the color labels
for different color regions, thus they are essentially semi-automatic. Aiming at
this drawback, in this paper we propose an automatic colorization technique
without any human interaction.
To automatically colorize a gray scale facial image, a natural idea is to reverse
the conversion of color to gray. A lot of recent work[7][8][9] focus on how to keep
information as much as possible during color image to gray scale conversion,
and several of them achieve remarkable results, such as spatial color to gray
for preserving chrominance edge information by Raja[7], and salience preserving
color to gray method by Amy A. Gooch[8], etc.
However, all those conversion methods above do not consider inverse process,
i.e. gray scale to color, and conversion from color to gray is usually irreversible
because the data dimension is reduced from 3 to 1 with huge information lost.
Thus converting a gray scale image to color one is essentially the problem of
recovering data from deficient data sample and the recovery problem can be for-
mulated as a probability to measure the recovery quality. Hence, our approach is
driven by predicting the color of the gray scale images based on MAP framework,
and our goal is to maximize the probability of the prediction. Furthermore, for
facial images a certain local geometrical configuration can be found to enhance
the recovery results.
To guarantee that the colorized facial image is acceptable, three criteria are
proposed as follows:
– Identity invariant: the recovered color face should look like the given gray
face.
– Global constraint: the common facial features should be kept during re-
covery process, which means the skin color, lip luster or other components’
color should be natural.
剩余10页未读,继续阅读
资源评论
weixin_38564826
- 粉丝: 5
- 资源: 910
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功