Professional Documents
Culture Documents
Textbook Artificial Intelligence and Algorithms in Intelligent Systems Radek Silhavy Ebook All Chapter PDF
Textbook Artificial Intelligence and Algorithms in Intelligent Systems Radek Silhavy Ebook All Chapter PDF
https://textbookfull.com/product/cybernetics-and-algorithms-in-
intelligent-systems-radek-silhavy/
https://textbookfull.com/product/software-engineering-and-
algorithms-in-intelligent-systems-radek-silhavy/
https://textbookfull.com/product/artificial-intelligence-methods-
in-intelligent-algorithms-proceedings-of-8th-computer-science-on-
line-conference-2019-vol-2-radek-silhavy/
https://textbookfull.com/product/computational-and-statistical-
methods-in-intelligent-systems-radek-silhavy/
Intelligent Systems in Cybernetics and Automation
Control Theory Radek Silhavy
https://textbookfull.com/product/intelligent-systems-in-
cybernetics-and-automation-control-theory-radek-silhavy/
https://textbookfull.com/product/software-engineering-
perspectives-in-intelligent-systems-proceedings-of-4th-
computational-methods-in-systems-and-software-2020-vol-1-radek-
silhavy/
https://textbookfull.com/product/software-engineering-
perspectives-in-intelligent-systems-proceedings-of-4th-
computational-methods-in-systems-and-software-2020-vol-2-radek-
silhavy/
https://textbookfull.com/product/intelligent-algorithms-in-
software-engineering-proceedings-of-the-9th-computer-science-on-
line-conference-2020-volume-1-radek-silhavy/
Artificial
Intelligence and
Algorithms in
Intelligent Systems
Proceedings of 7th Computer Science
On-line Conference 2018, Volume 2
Advances in Intelligent Systems and Computing
Volume 764
Series editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
e-mail: kacprzyk@ibspan.waw.pl
The series “Advances in Intelligent Systems and Computing” contains publications on theory,
applications, and design methods of Intelligent Systems and Intelligent Computing. Virtually all
disciplines such as engineering, natural sciences, computer and information science, ICT, economics,
business, e-commerce, environment, healthcare, life science are covered. The list of topics spans all the
areas of modern intelligent systems and computing such as: computational intelligence, soft computing
including neural networks, fuzzy systems, evolutionary computing and the fusion of these paradigms,
social intelligence, ambient intelligence, computational neuroscience, artificial life, virtual worlds and
society, cognitive science and systems, Perception and Vision, DNA and immune based systems,
self-organizing and adaptive systems, e-Learning and teaching, human-centered and human-centric
computing, recommender systems, intelligent control, robotics and mechatronics including
human-machine teaming, knowledge-based paradigms, learning paradigms, machine ethics, intelligent
data analysis, knowledge management, intelligent agents, intelligent decision making and support,
intelligent network security, trust management, interactive entertainment, Web intelligence and multimedia.
The publications within “Advances in Intelligent Systems and Computing” are primarily proceedings
of important conferences, symposia and congresses. They cover significant recent developments in the
field, both of a foundational and applicable character. An important characteristic feature of the series is
the short publication time and world-wide distribution. This permits a rapid and broad dissemination of
research results.
Advisory Board
Chairman
Nikhil R. Pal, Indian Statistical Institute, Kolkata, India
e-mail: nikhil@isical.ac.in
Members
Rafael Bello Perez, Universidad Central “Marta Abreu” de Las Villas, Santa Clara, Cuba
e-mail: rbellop@uclv.edu.cu
Emilio S. Corchado, University of Salamanca, Salamanca, Spain
e-mail: escorchado@usal.es
Hani Hagras, University of Essex, Colchester, UK
e-mail: hani@essex.ac.uk
László T. Kóczy, Széchenyi István University, Győr, Hungary
e-mail: koczy@sze.hu
Vladik Kreinovich, University of Texas at El Paso, El Paso, USA
e-mail: vladik@utep.edu
Chin-Teng Lin, National Chiao Tung University, Hsinchu, Taiwan
e-mail: ctlin@mail.nctu.edu.tw
Jie Lu, University of Technology, Sydney, Australia
e-mail: Jie.Lu@uts.edu.au
Patricia Melin, Tijuana Institute of Technology, Tijuana, Mexico
e-mail: epmelin@hafsamx.org
Nadia Nedjah, State University of Rio de Janeiro, Rio de Janeiro, Brazil
e-mail: nadia@eng.uerj.br
Ngoc Thanh Nguyen, Wroclaw University of Technology, Wroclaw, Poland
e-mail: Ngoc-Thanh.Nguyen@pwr.edu.pl
Jun Wang, The Chinese University of Hong Kong, Shatin, Hong Kong
e-mail: jwang@mae.cuhk.edu.hk
Artificial Intelligence
and Algorithms in Intelligent
Systems
Proceedings of 7th Computer Science On-line
Conference 2018, Volume 2
123
Editor
Radek Silhavy
Faculty of Applied Informatics
Tomas Bata University in Zlín
Zlín
Czech Republic
This Springer imprint is published by the registered company Springer International Publishing AG
part of Springer Nature
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
This book constitutes the refereed proceedings of the modern trends and approaches
of artificial intelligence research and its application to intelligent systems. The paper
discusses hybridization of algorithms, new trends in neural networks, optimization
algorithms, and real-life issues related to artificial methods’ application.
This book constitutes the refereed proceedings of the Artificial Intelligence and
Algorithms in Intelligent Systems of the 7th Computer Science On-line Conference
2018 (CSOC 2018), held online in April 2018.
CSOC 2018 has received (all sections) 265 submissions, 141 of them were
accepted for publication. More than 60% of accepted submissions were received
from Europe, 30% from Asia, 5% from Africa, and 5% from America. Researchers
from 30 countries participated in CSOC 2018 conference.
CSOC 2018 conference intends to provide an international forum for the dis-
cussion of the latest high-quality research results in all areas related to Computer
Science. The addressed topics are the theoretical aspects and applications of
Computer Science, Artificial Intelligences, Cybernetics, Automation Control
Theory and Software Engineering.
Computer Science On-line Conference is held online and modern communica-
tion technology, which is broadly used to improve the traditional concept of sci-
entific conferences. It brings equal opportunity to participate to all researchers
around the world.
I believe that you will find following proceedings interesting and useful for your
own research work.
v
Organization
Program Committee
Program Committee Chairs
vii
viii Organization
OpenPublish.eu s.r.o.
Web: http://www.openpublish.eu
Email: csoc@openpublish.eu
http://www.openpublish.eu
Contents
xi
xii Contents
1 Introduction
which integrate the situational context of an event (see [4]). The dimensional models
suggest that emotion can be explained by a dimensional space, such as a 2D space
spanning valence and arousal. Valence describes the extent to which an emotion is
positive or negative, whereas arousal refers to emotional intensity, i.e., the strength of
the associated emotional state.
Since advertising seeks to influence the user’s mental functions - first of all, per-
ception, behavior, thinking and emotions, the interest in the role of affective factors on
the perception of advertising is steadily growing, and so there is a growing number of
studies that help to understand how advertising is most affected by emotion [5, 6].
The task is made more difficult by known fact that the content can not be applied
indiscriminately, e.g., if one ad engages children, it will not necessarily be suitable for
adults. Moreover, there is a dual perception of the emotions visible in the image [7] as
follows:
(1) Syncretic cognition, which is occurring instinctively, spontaneously and broadly
subjectively. The image is perceived by what it is visually recognizable and
known from before, it causes relevant emotions (e.g., joy, sorrow, fear, etc.); and
(2) Analytical cognition, which is slower. The image is perceived by thinking about
the visible characters, analyzing their meanings. The analytical perception is less
subjective, since other well-known descriptions of the image are used.
While advertisers are often trying to disturb people emotions and thus force them to
engage in an advertised product or product, a user who first encounters a positive
impression on a product becomes more influenced by emotions (syncretic perception)
than rational decisions (analytical perception).
Recently, the researchers began evaluating the neurological correlates of consumer
behavior using the electroencephalography (EEG), electromyography (EMG), galvanic
skin response (GSR) and gaze tracking technology in order to understand
marketing-relevant human behavior [8]. The quantitative data collected and properly
analyzed provides an advantage over the commonly used surveys and questionnaires,
which rely on the ability and willingness of the participants to honestly and accurately
report their attitudes and behaviors.
Eye movements, gaze landing sites and areas of interest are known to be good
predictors of customer behavior. People spend more time for examining (i.e., fixating
their gaze on) options that they will eventually choose [9]. For example, consumers
spent 54% more time looking at the ads of businesses that they ended up choosing [10].
Unusual brand extension cause disruption of visual processing and result in longer
visual fixation times as evidenced in [11]. The emotional state can be recognized using
gaze tracking technology [12], while different factors such as the emotional state and
gender have influence on visual processing and understanding [13].
The aim of the paper is to analyze gender and affect related differences in the
perception of visual advertising. The results of experiments using gaze tracking are
presented and analyzed, and the implications for neuromarketing are discussed.
The structure of the remaining parts of the paper is as follows. Section 2 describes
the experiments. Section 3 evaluates the results and concludes.
Analysis of Affective and Gender Factors in Image Comprehension 3
2 Experiments
2.1 Subjects
The experiment was conducted at Kaunas University of Technology (Kaunas,
Lithuania). Due to the specifics of the hardware requirements, the place of execution
and the limited access to the gaze tracking device, only students of 18–25 years old
were invited to participate. As this social group of 18–25 age range is often a target
audience in commercial advertising, we consider the use of this social group in research
as appropriate [14]. Twenty people - 10 women (50%) and 10 men (50%) have par-
ticipated. 45% of respondents were 18–21 years old, while 55% were 22–25 years old.
All subjects reported to have normal vision. Participants signed a written consent form.
The participants also received written task instructions on the procedure.
2.2 Dataset
The analyzed data included three original ads and their modified versions with emo-
tions introduced (changed) using image processing software as follows (also see
Fig. 1): “Trump for President 2016” (source: http://adsoftheworld.com/), “Charming
Chocolate”, “Charming Chocolate” ad created by the students of KTU and “I am
Muslim” ad by People Against Suffering Oppression and Poverty (PASSOP).
Fig. 1. Analyzed images: original (top) and modified (bottom): “Trump for President 2016”
(left), “Charming Chocolate” (center), “I am Muslim” (right)
4 G. Liaudanskaitė et al.
2.3 Methodology
The experiment was performed on the basis of a mixed quantitative and qualitative
research methodology. To achieve the objective, an eye tracking device and special
software were used. Research was conducted using the methodology presented in [15],
which sought to find out how people react to smiling faces in promotional imagery. The
aim is to find out whether two groups of subjects of similar composition, which are
shown the same advertisements in which only the expression of a person’s face differs,
can notice the difference in the intensity of the reaction.
For each participant the gaze-tracking device was adjusted and calibrated. Each
experiment consisted of 3 stages, while calibration was repeated at the beginning of
each stage. Following the experiments, the raw data was pre-processed and eye fixa-
tions shorter than 60 ms were discarded because of contamination from equipment
measurement noise and blinking effects [16]. The upper threshold of fixations has been
set to 600 ms as recommended in [17].
2.4 Procedure
Before the experiment, the intensity of human emotions seen in each advertisement was
measured using FaceReader software (Noldus, the Netherlands). The program can
recognize the basic facial expressions: happiness, surprise, sadness, anger, fear, con-
tempt, disgust, and neutrality. The intensity of each emotion is evaluated on a scale
from 0 to 1, where zero means that emotion is not recognized, and one - the emotion is
very strong. The results of emotions measurements are shown in Table 1.
The participants were divided into two equal groups with an equal number of
women and men of the same age, emotional advertisements were displayed, or mod-
ified with the use of additional photos for retouching. One ad was displayed for a
maximum of 30 s, but each respondent was able to switch the ad on when she did not
want to watch it for longer. Numeric data collected during the experiment was the
timeframe for fixation of sight of the software (start time and duration in ms).
Analysis of Affective and Gender Factors in Image Comprehension 5
k x k1 ðx=kÞk
pdf ð xÞ ¼ e ð1Þ
k k
where k > 0 is the shape parameter and k > 0 is the scale parameter of the distribution.
A value of k > 1 indicates that the failure rate increases with time. This happens if there
is an aging or fatigue [20] process.
Weibull distribution parameters, gamma and beta, highly correlate (between 84 and
93%) with neural responses in the early visual system [21]. Sometimes none of the
considered models represents the data appropriately. Thus, it should be tested how well
the integrated Weibull distribution fits the data. The goodness of fit between experi-
mentally established distribution and real data is evaluated using R2 value, which is
calculated as follows:
Pn
ð^xi xÞ2
R2 ¼ Pi¼1
n
ð2Þ
i¼1 ðxi xÞ 2
and
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 1 2
std ð X Þ ¼ k C 1 þ C 1þ ð4Þ
k k
To evaluate if two data samples belong to the same statistical distribution, a number
of statistical tests can be used.
The Kruskal-Wallis method tests if two or more classes have equal median and
gives the value of p. If the value of p is close to 0 it means that the feature contains
discriminative information.
6 G. Liaudanskaitė et al.
The Wilcoxon rank sum test is used to establish if there is a significant difference
between two sample groups using their ranks. If the ranks of the sample groups are
significantly ða ¼ 0:05Þ separated, the test statistic identifies significant difference.
The third test statistic is obtained by bootstrapping, a way of estimating statistical
parameters from the sample by performing resampling with replacement. For a
two-sample bootstrap, we independently draw bootstrap samples with replacement
from each sample, and compute the Fisher–Pitman test [22] statistic that compares the
samples. The bootstrap distribution is used for confidence intervals and standard errors.
Given the observed value of the difference (D) between the means of the samples
x ¼ ðx1 ; x2 ; . . .; xn Þ and y ¼ ðy1 ; y2 ; . . .; yn Þ drawn from two probability distributions X
and Y, the p-value is calculated as follows:
PB
1þ ½Di jDj
pvalue ¼ i¼1
ð6Þ
Bþ1
here ½ is the Iverson bracket operator, and B is the number of bootstrap resamples, here
we set it to 1000 as suggested in [23].
For calculation of confidence, the percentile intervals are used. The sample values
are arranged in ascending order and numbered sequentially with the smallest value
receiving the number 1, obtaining the ranks of the sample values. For 95% confidence
interval, the 2.5 and 97.5 percentiles are used, which corresponds to 0:025n þ 0:5 and
0:975n þ 0:5 ordered values, where n is the sample size [24].
2.6 Hypotheses
H1. There is no difference between genders in eye fixation times. Hypotheses are:
ð1Þ ð1Þ
H0 : lmale ¼ lfemale ; H1 : lmale 6¼ lfemale
ð2Þ ð2Þ
H0 : lhappy ¼ lsad ; H1 : lhappy 6¼ lsad
ð1aÞ ð1aÞ
H0 : lhappy;male ¼ lhappy;female ; H1 : lhappy;male 6¼ lhappy;female
Analysis of Affective and Gender Factors in Image Comprehension 7
H1b. There is no difference between genders in eye fixation times of the negatively
valenced images. Hypotheses are:
ð1bÞ ð1bÞ
H0 : lsad;male ¼ lsad;female ; H1 : lsad;male 6¼ lsad;female
H2a. There is no difference between affect of images in the eye fixation times in
males. Hypotheses are:
ð2aÞ ð2aÞ
H0 : lhappy;male ¼ lsad;male ; H1 : lhappy;male 6¼ lsad;male
H2b. There is no difference between affect of images in the eye fixation times in
females. Hypotheses are:
ð2bÞ ð2bÞ
H0 : lhappy;female ¼ lsad;female ; H1 : lhappy;female 6¼ lsad;female
2.7 Results
All data were analyzed with custom written scripts and executed in MATLAB
(Mathworks, Inc., Natick, Massachusetts). First, we determine the statistical distribu-
tion of data by fitting. Following [25], Weibull distribution can be used as a simple
predictor of fixation location data, which can be explained by the image local contrast
statistics. In our case, the goodness-of-fit between real data and fitted Weibull distri-
bution data is 0.96.
When analyzing the eye fixation time data in the Weibull distribution parameter
space (scale parameter k vs shape parameter k, see Eq. 1), we can notice a difference
both in gender and affective factors (see Fig. 2). Here individual parameter values were
generated from initial population of data by bootstrapping. For males, the probability of
having a larger value of scale parameter k than for females is p = 0.6274, while for the
shape parameter k, the difference is not significant (p = 0.6191). In the affective
dimension, the probability of a larger value of the scale parameter k for happy vs sad
samples of data is p = 0.8493, while the probability of a larger value of the shape
parameter k is p = 0.9308. This means that positive emotions lead to longer mean
fixation times and more outliers in the right tail of the value distribution.
Following [26], we differentiate between short (correcting and ambient) fixations,
and long (focal) fixations. Long fixations last longer than 300 ms and are attributed to
conscious visual processing process, while shorter fixations are related to unconscious
processing of information.
8 G. Liaudanskaitė et al.
Fig. 2. Gaze data in the Weibull distribution parameter space (with 0.95 confidence ellipses)
In this paper, we have analyzed the affect and gender related differences of how
subjects react to emotionally charged advertisements. Four different statistical tests
(Kruskal-Wallis, ranksum, ANOVA, bootstrapping) confirmed the significant differ-
ence (p < 0.05) between short eye fixation times (<300 ms) in the affective dimension,
however we failed to discover any gender related differences of how the images are
processed by human visual system.
This observation agrees well with earlier findings in adult subjects, suggesting that
there is a direct link between eye movement behavior characterized by fixation duration
and distinct modes of visual processing for different visual stimuli (such as photographs
and emotional stimuli [27] as well as for different tasks (localization vs. identification
[28]). Specifically, the ambient mode characterized by short fixation times has been
related to pre-attentive scanning associated with the visual exploration of the spatial
layout of an image, which is characteristic to the early stages of viewing [28]. Our
findings are in line with these previous observations.
However, the findings of this study are preliminary and limited in nature. We need
to repeat the study with a larger dataset and a larger population of subjects to confirm or
reject the present results. Despite these potent limits, the results of the current study are
promising in terms of demonstrating a new modality, which can serve as an assistive
tool for evaluation of visual ads in the context of neuromarketing.
References
1. Clay, R.A.: Advertising as science. Am. Psychol. Assoc. 33(9), 38 (2002)
2. Senior, C., Lee, N.: A manifesto for neuromarketing science. J. Consum. Behav. 7(4–5),
263–271 (2008)
3. Levenson, R.W.: Basic emotion questions. Emotion Rev. 3, 379–386 (2011)
4. Ellsworth, P.C., Scherer, K.R.: Appraisal processes in emotion. In: Davidson, R.J.,
Goldsmith, H., Scherer, K.R. (eds.) Handbook of Affective Sciences, pp. 572–595. Oxford
University Press, New York and Oxford (2003)
5. Yadati, K., Katti, H., Kankanhalli, M.S.: CAVVA: computational affective video-in-video
advertising. IEEE Trans. Multimedia 16(1), 15–23 (2014)
6. Kwong, C.K., Jiang, H., Luo, X.G.: AI-based methodology of integrating affective design,
engineering, and marketing for defining design specifications of new products. Eng. Appl.
AI 47, 49–60 (2016)
7. Chaudhuri, A., Buck, R.: Media differences in rational and emotional responses to
advertising. J. Broadcast. Electron. Media 39(1), 109–125 (2009)
8. Lee, N., Broderick, A.J., Chamberlain, L.: What is ‘neuromarketing’? A discussion and
agenda for future research. Int. J. Psychophysiol. 63(2), 199–204 (2007)
9. Krajbich, I., Armel, C., Rangel, A.: Visual fixations and the computation and comparison of
value in simple choice. Nat. Neurosci. 13(10), 1292–1298 (2010)
10. Lohse, G.L.: Consumer eye movement patterns on yellow page advertising. J. Advertising
26, 61–73 (1997)
11. Stewart, A.J., Pickering, M.J., Sturt, P.: Using eye movements during reading as an implicit
measure of acceptability of brand extensions. Appl. Cognit. Psychol. 18, 697–709 (2004)
12. Maskeliunas, R., Raudonis, V.: Are you ashamed? Can a gaze tracker tell? PeerJ. Comput.
Sci. 2, e75 (2016)
13. Raudonis, V., Maskeliunas, R., Stankevicius, K., Damasevicius, R.: Gender, age, colour,
position and stress: how they influence attention at workplace? In: 17th International
Conference on Computational Science and Its Applications (ICCSA) vol. 5, pp. 248–264
(2017)
14. Coursaris, C.K., Sung, J., Swierenga. S.J.: Effects of message characteristics, age, and
gender on perceptions of mobile advertising – an empirical investigation among college
students. In: Ninth International Conference on Mobile Business and 2010 Ninth Global
Mobility Roundtable (ICMB-GMR), pp. 198–205 (2010)
15. Berg, H., Söderlund, M., Lindström, A.: Spreading joy: examining the effects of smiling
models on consumer joy and attitudes. J. Consum. Market. 32(6), 459–469 (2015)
16. Helo, A., Pannasch, S., Sirri, L., Rämä, P.: The maturation of eye movement behavior: scene
viewing characteristics in children and adults. Vis. Res. 103, 83–91 (2014)
17. Jacob, R.J.K.: Eye tracking in advanced interface design. In: Barfield, W., Furness, T. (eds.)
Virtual Environments and Advanced Interface Design. Oxford University Press, Oxford
(1994)
18. Bosworth, R.G., Petrich, J.A.F., Dobkins, K.R.: Effects of attention and laterality on motion
and orientation discrimination in deaf signers. Brain Cognit. 82, 117–126 (2013)
19. Yanulevskaya, V., Geusebroek, J.M.: Significance of the Weibull distribution and its
sub-models in natural image statistics. In: Proceedings of the International Conference on
Computer Vision Theory and Applications, pp. 355–362 (2009)
Analysis of Affective and Gender Factors in Image Comprehension 11
20. Vasiljevas, M., Gedminas, T., Sevcenko, A., Janciukas, M., Blazauskas, T., Damasevi-
cius R.: Modelling eye fatigue in gaze spelling task. In: 2016 IEEE 12th International
Conference on Intelligent Computer Communication and Processing (ICCP), pp. 95–102
(2016)
21. Scholte, H.S., Ghebreab, S., Waldorp, L., Smeulders, A.W.M., Lamme, V.A.F.: Brain
responses strongly correlate with Weibull image statistics when processing natural images.
J. Vis. 9, 1–9 (2009)
22. Fisher, R.A.: The Design of Experiments, 8th edn., pp. 1–248. Hafner Publishing Company,
New York (1971)
23. Briggs, A.H., Wonderling, D.E., Mooney, C.Z.: Pulling cost-effectiveness analysis up by its
bootstraps: a non-parametric approach to confidence interval estimation. Health Econ. 6,
327–340 (1997)
24. Diem, K., Seldrup, J., Lentner, C. (eds.): Geigy scientific tables. In: Introduction to Statistics,
Statistical Tables and Mathematical Formulae—Percentiles, vol. 2, 8th edn. Ciba-Geigy
Limited, Basel (1982)
25. Yanulevskaya, V., Marsman, J.B., Cornelissen, F., Geusebroek, J.-M.: An image
statistics-based model for fixation prediction. Cognit. Comput. 3(1), 94–104 (2011)
26. Velichkovsky, B.B., Rumyantsev, M.A., Morozov, M.A.: New solution to the midas touch
problem: identification of visual commands via extraction of focal fixations. Procedia
Comput. Sci. 39, 75–82 (2014)
27. Pannasch, S., Helmert, J.R., Roth, K., Walter, H.: Visual fixation durations and saccade
amplitudes: shifting relationship in a variety of conditions. J. Eye Mov. Res. 2(2), 1–19
(2008)
28. Unema, P., Pannasch, S., Joos, M., Velichkovsky, B.: Time course of information processing
during scene perception: The relationship between saccade amplitude and fixation duration.
Visual Cognit. 12(3), 473–494 (2005)
FARIP: Framework for Artifact Removal
for Image Processing Using JPEG
1 Introduction
Image forensic does increasingly gaining attention owe to its explicit capability to
disclosing the underlying fact [1]. The prime assumption in this concept is any oper-
ations on the top of the image will statistically alter the significant metadata information
of an image [2]. Such statistical unique characteristics are called as fingerprint. The
validation process includes identification of such digital fingerprints also called as
traces or artifacts [3]. At present, there has been various research work being carried out
towards identification/localization of such traces [4, 5]. In majority of the existing cases
of research, there is a dependency on a large section of the signal in order to carry out
the analysis. Image compression plays a significant role in this regards. There is a
higher significance level for the fingerprints generated as majority of the images are
subjected to the compression by their capturing device or during the process of storing
it in memory [6]. When an attacker tampers the image, they make use of various
forgery tools in order to invoke an attack [7] by ensuring that both the tamper image
and original image doesn’t bear a strong perceptual difference. The adversary has
multiple ways to compute such difference in order to avoid getting caught and hence it
is essential that a unique compression technique be explored in order to ensure proper
2 Related Work
This section upgrades the information about research contribution following our prior
review work [10]. Identification of traces in images has been investigated by Pasquini
et al. [11] using statistical approach. Lipman et al. [12] have addressed the reflection
artifacts for enhancing image quality on two dimensional shear waves. Adoption of
sparse representation along with structural quantization approach was studied by Zhao
et al. [13]. Liu et al. [14] improved the performing of deblocking technique using
knowledge-based approach. Study towards elimination of artifacts using filtering
mechanism has been seen in the work carried out by Gong et al. [15]. Fan et al. [16]
have performed an investigation towards spatial and transformed domain and its effect
on artifacts removal in images. Ebrahimi et al. [17] have presented a passive technique
for identification of blocking artifacts using experimental approach. Min et al. [18]
have studied the possible impact of artifacts arise from compression on perceptual
quality. Study towards nano-scale imaging along with its artifacts minimization was
carried out by Fairbairn and Moheimani [19]. Adoption of predictive approach is being
carried out by Bannan et al. [20]. Li et al. [21] have introduced a solution towards
artifacts removal from compression process to be used in steganalysis process. Study
using case study of medical image and its associated artifacts effect has been investi-
gated by Mauldin et al. [22] by analyzing geometric factors of the image. Sharma et al.
[23] have presented a study where both image quality and compression performance is
equally emphasized. Study towards localizing the image forgery has been carried out
by Bianchi et al. [24] considering the artifacts caused due to JPEG compression.
Fairbairn [25] has implemented a controller design for reducing the image artifacts.
Ferrara et al. [26] have carried out the study to localize the region of image forgery
considering probability map of artifacts. Goto et al. [27] have presented a study for
minimizing the compression artifacts using total variation technique. Johnson et al. [28]
have presented visual analysis technique for investigating the compression artifacts on
pathological images. Similar technique has been also carried out by Kim and Sim [29]
using adaptive approach based on signal weight. Tang et al. [30] have presented
Markov model for eliminating blocking artifacts using indexing mechanism. The next
section outlines the research problems.
14 T. M. Shashidhar and K. B. Ramesh
3 Problem Description
A closer look into the existing techniques on artifact removal shows that they are quite
capable of addressing the visible traces associated with blocking. Such models are
restricted to images of minimal to medium scale quality. However, for better perfor-
mance, there is a need for eliminating both visual as well as traces from statistical
approaches. Such process must not result in an image that could be detected forensi-
cally incorporated fingerprints. Moreover, a very simple technique is also demanded
using JPEG-based compression approach that could offer better computational time and
works on multiple forms of images. The next section introduces the proposed solution
to address this problem.
4 Proposed Solution
The proposed system of FARIP uses a very simple modeling of compression which
also performs removal of the traces. The proposed solution of addressing the problems
associated with the blocking artifacts is based on the fact that JPEG compression could
be suitably modified in such a way that all the traces could be significantly eliminated
precisely. The scheme of proposed system is showcased in Fig. 1.
Figure 1 highlights how a normal image is inflicted with blocking artifacts and
slowly the proposed FARIP performs removal of any form of traces during the
blocking process resulting in final decompressed image with good signal quality. The
complete process of artifacts removal is carried out using JPEG standard emphasizing
on the quantization steps. The next section outlines adopted research methodology. In
the above Fig. 1, the original image (Io) will be selected from the desired folder and its
size is computed. Later, a compressed image (Ic) will be obtained from Io.
The difference among both the Io and Ic with respect to PSNR and MSE and finally, the
FARIP: Framework for Artifact Removal for Image Processing 15
performance analysis is performed on the results obtained from existing method and
proposed method. The performance analysis is done by considering the image size and
compression quality parameters to explore the significance of proposed method.
The proposed FARIP is subjected to the artifact removal using the JPEG standard itself.
For this purpose, we apply Laplace distribution for initiating the system design con-
sidering the distribution of the coefficients over a range of specific bands as follows,
In the above Eq. (1), a represents parametric model and where A is transform
coefficient. The constant c represents c/2, where c represents unique value of the
sub-band. The exponential power variable d will represent – c|a| The distribution of the
priorly compressed JPEG image is carried out using discrete approach of Laplace
distribution as follows,
8 9
< 1 ed 1 b ¼ 0 =
aðB ¼ bÞ ¼ ed2 b ¼ kbi;j ð2Þ
: ;
0 otherwise
The exponential term d1 and d2 will mean c.bi,j/2 and c|b| respectively. In order to
perform any form of modification of the priorly compressed image, the initial operations
of JPEG compression steps were carried out in order to acquire transformed coefficients.
It is due to the fact that the ultimate phase of the decompression using JPEG consists of
projecting the decompressed values of the pixels back to coefficient values acquired
from the quantized image. Hence, it is feasible to obtain the coefficient values by
iterating the process of quantization on the scale of new coefficient B′ in such a way that
B = bi,j (B′/bi,j). The similar quantized coefficients were obtained using a highest
probability value of model attribute. With an aid of this process, it is now feasible to
compute the distribution of sub-band coefficient prior to performing JPEG compression.
We consider that η represents cumulative observations from all the sub-bands where η is
sum of number ηo of sub-band coefficient (that is usually valued as zeros) and all the
non-zero coefficient η1. Therefore, the empirical expression will be,
X
g
d¼ jbk j ð3Þ
k¼1
Once the c is computed, the anti-forensic dither is added for the entire coefficient in the
sub-band. In majority of the records of quantization, the sub-bands with higher coefficient
is used as it doesn’t offer higher significant perceptual changes. It has also been seen that
sub-bands and their respective variance minimizes when there is a sub-band transition
from lower to higher frequency. Owing to this process, the different forms of sub-band that
has higher value of coefficient is then subjected to quantization to the scale of zero while
Another random document with
no related content on Scribd:
only flounder through and up to the top. Besides, it was necessary to wade
the creeks. This wet my feet and caused more snow to bunch and freeze on
them. Night came, and with it an increase of cold, which, causing the snow
to freeze to a crust on top, iced and smoothed the track anew for me. But
with one additional facility for making progress, I lacked another. That was
the strength and freshness with which I had started at morn.
My day had been one of most laborious progress, wading creeks,
floundering through their soft snowbanks, and stopping every ten minutes
to clean my shoes of damp snow. I had no other grease for their bottoms
save a bit of pork, which I wore out upon them. Snowshoes won’t run well
unless frequently greased. Then there was no rest for my body. The supply
of pines from about whose roots the snow had melted away had given out;
to step off the shoes was to sink to the middle; to rest at all was to rest
squatting; a few minutes’ trial of this position under the most favorable
circumstances will convince any reader of its back-aching tendencies. Man
is a lying animal; I mean he must lie down to recuperate. My meals I
cooked on the snow. The regular menu was coffee, bread, and pork. The
base of the kitchen was a big piece of dry pine bark, always at hand. On this
the fire was kindled. The evolution of coffee under these conditions was
slow, because the water for making it had first to be melted from snow in
the coffee-pot, and snow under these circumstances melts with an
exasperating slowness. The quantity required to make a single pint of water
is something remarkable. I think I was obliged to fill that vessel four or five
times with snow to get the suitable quantity of water. Then it must be
remembered that the water would not proceed to boil until all the snow was
melted. “A watched pot never boils,” but a watched pot of plain water is
velocity itself when compared with a pot of snow and water watched by a
tired and hungry being in the wilderness. Night found me twelve miles from
Strawberry Flat. About one in the morning I found another empty cabin.
This was four miles from the desired haven. It was desolation inside and out
—the windows gone, the door torn from its hinges, the inside a litter of
snow and rubbish, and one dead cow in the kitchen. The cooking-stove
remained. I cleared the grate of snow and attempted a fire. It wouldn’t draw.
Of course it wouldn’t draw, for the stovepipe was full of snow. Then I
kindled the fire on the top of the stove and gradually burned up a portion of
the house. Like Sherman and Napoleon, I lived on the country invaded. The
firelight cast its ruddy glow on the surrounding domestic desolation and the
red dead cow, which, being frozen hard as a rock, served for a seat. I waited
for the morn; but the morn would not come. I saw from the sashless
windows the preliminary streaks of dawn ever so faintly lighting up the
eastern horizon full forty times, and found they were only in my
imagination, so covetous for the coming day. When the sun did rise he came
up in the opposite direction.
Impatient of waiting longer, and converting myself to the false belief that
the light was really coming in what turned out to be the west, I did battle in
the dark with the last four miles of the journey. All went well until I reached
a certain point in the road, which was now well defined through the trees.
There every time I brought up in a clump of bushes and lost the track. Back,
time after time, I went, seeking with careful calculation to make a fresh and
truer start, only to bring up again in brambles and briers. When the light did
come, I had, for two hundred yards, in this going and coming, beaten a path
in the snow which looked as if travelled over for a week. Daylight showed
what a ridiculously trifling turn had caused me thus to miss the route. The
moral of which is that man’s welfare is often wrecked on some trifling
error. A failure to say, “Good morning” to an acquaintance, a long, gloomy
countenance or putting one’s knife in the mouth, or guzzling down soup or
coffee with undue noise, has repelled one man from another, and such
repulsion has sent our fortunes on the wrong track altogether. I was
welcomed at the Strawberry House by six hounds, who, in the still faint
light, made at me as they would at another beast, and the first few moments
of my arrival on this outpost of civilization was occupied in energetic
attempts to keep what was left of me from being eaten by the dogs. Indeed,
the way of the transgressor is hard. I had never injured these dogs.
However, the hospitality I experienced at Palmer’s stood out in bold relief
against the churlishness of their brute creation.
CHAPTER XXIX.
ON THE ROSTRUM.
Previous to this election which did not elect me, Williams and I
canvassed the county together. He aspired to the office of Sheriff. We
mounted our horses, and with long linen dusters on our backs and bottles of
whiskey in our pockets, rode first to Spring Gulch, consisting of two
groceries, six saloons, an empty hotel, twenty miners’ cabins, a seedy
school-house, a seedier church, the hillsides around denuded of earth, torn
and scarred by years of hydraulic washing, and showing great patches of
bare yellow ledge covered with heaps of boulders. The few men met were
in coarse, ragged, gray shirts and mud-stained duck pants, had a worn,
worked-out look; over all shining the hot afternoon sun, the heated
atmosphere quivering and rising, behind the hill-bounded horizon, a snow-
white mass of cloud which, at precisely the same hour every afternoon,
attains the same altitude, then gradually sinks. The eye gazes steadily upon
it; there are seen great hollows and depths of shining whiteness. It is the
vapor coming from the melting snow on the Sierra peaks eighty miles away.
The few loungers about the Washington Saloon see William Saunders and
myself riding down the hill. Our dusters and clean linen proclaim us as
“candidates.” Candidates means drinks. There is a gradual concentration of
unemployed seediness at the Washington. We dismount; soon the coveted
and cheering bottle is placed on the bar; a line of tumblers in skirmishing
order form behind it; every one within sight and hearing is called up; a
pause of glad anticipation ensues while the glasses are being filled; the
precision of bar-room etiquette is strictly observed, that not a drop be
swallowed until all are ready; then the dozen tumblers are simultaneously
raised; the standing toast “Here’s luck,” and the reviving alcohol fulfils its
mission. This is electioneering.
Sam White is the Bismarck of our interests in Spring Gulch. He is the
standing delegate to the County Convention from this precinct. He goes by
virtue of a paying claim, a capacity for venturing among the rocks and
shoals of saloons, gaming tables and innumerable calls to drink, without
losing his head. He can drink deeply, quietly, and fearfully; he can drink
himself into noise and turbulence and still keep a set of sober faculties in
reserve underneath. We hold a short cabinet meeting with Sam behind the
barn. He sees clearly the political complexion of Spring Gulch. Bob
O’Leary is doubtful, but may be bought; Jack Shear and Tom Mead must be
braced up to allegiance by whiskey; Miles and O’Gorman are mad because
a favorite of theirs could not get the nomination for Supervisor last year,
and won’t vote anyhow; Bob Jones is favorable to us, but wants to leave
before the primary meeting comes off; the rest are sure for us or sure
against us.
We visit the Franklin House just opposite. The political candidate’s
money must not all be spent in one house. This is one of the fundamental
principles in electioneering. Every saloon controls a few votes, or rather a
few whiskey-sodden organizations, who are voted like machines. The
solemn ordeal of an American treat is again witnessed. Jim Brown becomes
affectionately and patriotically drunk, and as we ride away loudly proclaims
himself a “white man and in favor of a white man’s government.”
We feel that Spring Gulch is secure. We carry it in our pocket. We ride a
couple of miles over the ridge to Six-Bit Gulch. Red crags tower upward for
hundreds of feet; a rivulet flows along, and on a little flat under a spreading
live-oak is an old log cabin. In front is a bit of vegetable garden inclosed by
old sluice lumber. High up in the branches overhead a gauze-covered meat-
safe; on the trunk is nailed a coffee-mill; under it hangs a frying-pan; close
by the washtub and wash-board a few fowl peck about; the quail in a clump
of chaparral near by are querously twittering, scolding and fluttering, and
making the preliminary arrangements for their night’s rest.
Sam Lugar, gray and worn, resident in this gulch for the last sixteen
years, sits outside the door smoking his evening pipe.
A hundred yards above is the residence of the “Judge,” another hard-
working, whiskey-drinking hermit. A glance within shows the Judge eating
his evening meal. A child is playing about on the mud floor, whose creamy
complexion and bright bead-like eyes indicate its Indian origin. Hanging
above the fire-place are a gun, an Indian bow, a quiver full of glass-tipped
arrows; on the shelf bits of gold-studded quartz, a bunch of crystals,
petrifactions, and curiously-shaped stones found by the “Judge” from time
to time in his diggings. There are boxes full of old magazines and
newspapers; on the rude window-sill a coverless, well-worn copy of
Shakespeare. The Judge is tall, straight, and sallow in complexion. He has
lived on this spot since 1849. Six-Bit Gulch was very rich. He has torn up
virgin gold in the grass roots. He lives now on recollections of the flush
times. Present failures and long past successes form the staple of his
conversation. His mining is merely secondary to another occupation, the
great aspiration of his life—to beat a poker game over in Spring Gulch. He
has been unsuccessfully trying this for the last seven years. A bundle of
aboriginal duskiness enveloped in a bright calico gown, hanging about her
adipose proportions, stirs as we enter. That is the Judge’s wife—a squaw.
Her family down to the third generation, are camped in the brush hard by.
They visit the Judge at stated intervals, and at such times the family
expenses are trebled. The gray shirt and duck pants tied at the waist with a
string constitute the Judge’s only dress-suit. On the floor near him is a
shapeless, wet mass of India rubber boots, shirt and pants, drenched and
splashed with yellow mud. This man was once a spruce clerk in a New
England store. At seventeen, the set and whiteness of his collars, the fit of
his boots, the arrangement of hair and neck-tie were subjects of long and
painful consideration before the mirror. He had his chosen one among the
village girls; he saw her regularly home from the Sunday-evening prayer-
meetings. The great gold fever of 1848 seized him. He saw a vision: A few
months picking up nuggets in California; a triumphant return home; a
wedding; a stylish mansion; a fast horse; a front pew; termination, a marble
monument in the Terryville cemetery: “Beloved and respected by all who
knew him, he sleeps in hope of a still brighter immortality.”
We stop at the “Judge’s” for the night. Wife and child are sent off to the
Indian camp in the chaparral. Sam Lugar drops in after supper. The Judge is
an incessant talker. The bottles and glasses are placed on the table. The
Judge becomes fatherly as to counsel and admonition against excess in
drink. Also against gambling. He has peculiar theological views. Moses, he
says, was a keen old miner. He and Aaron put up a plan to gain all the gold
in the Israelites’ possession. While Moses was on Mount Sinai receiving the
stone tables, Aaron was counselling the making and worship of the golden
calf. By such means did he concentrate in a lump all the Jews’ jewelry.
What then? Moses comes down, sees the calf, gets angry, breaks into
pieces, burns it up. But what becomes of the gold? Didn’t Moses and Aaron
sneak around that night and “pan it out” of the ashes?
The Judge is his own theologian.
We visit Price, of Hawkins’ Bar. Price is now the sole constituency of
Hawkins’. He ran this bar in its golden infancy; he saw it in its youth; he is
steadfast to it in its decay. Thirty-four years ago, eight hundred men lived
here; the Tuolumne banks were lined with them, shaking their cradles.
From the top of yonder red hill the combined grating of the pebbles shaken
in hundreds of rocker-sieves sounded like the crash of machinery in a cotton
mill.
Old Hawkins first discovered gold here. Price tells of the pickle-jars full
he had buried under the floor of his cabin. The secret could not be kept.
They came trooping down the steep Red Mountain trail, blankets and tools
on their backs, footsore, weary, thirsty, hungry—but hungrier still for gold.
They put up tents and brush houses, or crept, slept and cooked under
projecting rocks; they stood all day in ice-cold water; they overworked
bodies hitherto unused to manual labor; they blistered delicate hands; they
lived on bacon and heavy bread of their own making; they drank raw
whiskey by the quart; they died, and were buried almost where they died, in
nameless graves. Up yonder, but a few yards in the rear of Price’s cabin, is
the old camp graveyard. The fence is rotting away and stands at various
angles. The inscriptions on the headboards are half effaced by time and the
elements. Some are split and have fallen down. Read “Jacob Peiser, æt. 27.”
He died close by in the gulch hard by, with a pistol-bullet through him. A
dispute over a claim. “Samuel Purdy, 31.” Drowned trying to cross the river
during a freshet. “John Wilkins, æt. 35.” Killed by a cave in the bank claim
about a hundred yards away. “Samuel Johnson, æt. 25.” He dove with a
sand bag to stop a great leak in the Ford Chann’s head wall, and he stopped
the leak in part with his own body, for the stream sucked him in the crevice
and he never came up alive. “John Weddell, 35.” Blown up by the
premature explosion of a blast in the Split Rock quartz claim. “Abram
Hewison, 45.” Delirium tremens, stark mad at midnight, jumped into the
river from the point yonder, where the stream whirls round the bend with
tremendous force and then rushes down toward the long deep cañon a mile
away in a succession of great white crested billows, whose sad, never-
ceasing murmur seems an eternal requiem for those lying here.
Price has seen all this. That was the climax of his life. Price’s heaven is
not in the future. It is in the past. It is embraced in a period about twenty-
five years ago, when he made “an ounce per day.” Those, he remarks, were
times worth living for. Eight hundred souls then at Hawkins’; five gambling
houses in full blast every night; music, dancing, and fandangos at either end
of the bar.
The river roars unvexed toward the sea. It has burst through its dams and
choked the races with sand. The scars and furrows on the hill sides are quite
hidden by the thickly growing vegetation; young oaks and pines are coming
up in the place of the old. Trail and road are overgrown with brush. Among
the rank weeds we stumble on traces of man’s former presence—the top of
a saloon counter, the mahogany leg and faded green cushion of a billiard
table, rusty tin ware, broken picks and shovels, a few rude stone chimneys,
about whose blackened fire-places years ago gathered the hopeful, sanguine
men of “ ’49.” It is so still. The declining afternoon sun is throwing long
shadows from the mountains on the other bank. Slowly they creep up and
shade the steeps on our side. Every moan and babble of the Tuolumne falls
distinctly on the ear.
“Civilization” here put in a transient appearance. It scarred the hill sides
with pits and furrows dug for gold. It cut down the wide-spreading
symmetrical oaks. It forced the Tuolumne through race and flume from its
channels. It built gaudy temples dedicated to the worship of Bacchus,
resplendent with mirrors, pictures, and cut-glassware, located on the very
site where a few months previous stood the Indian’s smoking wigwam. It
brought toiling men, hard-fisted, awkward, ungainly, clumsy, with all grace
and suppleness worked out of them and strong only to lift and dig. It
brought all manner of men, educated and ignorant, cultivated and coarse,
yet for whom Christian training, Christian Church, Christian Bible,
Christian spire in city, town, and village pointing heavenward, had failed to
convince that gold was not the chief aim and end of all human effort. By
day there was labor drudging, labor spasmodic, a few prizes, many blanks,
some hope, much more discouragement. By night, revelry, carousal,
gambling, oaths, recklessness, pistol shots, knife thrusts, bloodshed, death.
Bird and beast fled affrighted to lonelier and more secure retreats before the
advent of the raging, cruel animal man.
But now civilization has flown and nature seems easier and somewhat
improved by its absence. Price is ours. He will walk nine miles on election
day to Chinese Camp, the nearest precinct, to deposit a ballot for us. An
order on the proprietor of the Phœnix Saloon for a generous supply of
whiskey stimulates his devotion to his country. What a glorious land of
liberty is this! See in the clear azure sky above us, floating a mere speck,
the eagle, the bird of freedom! He poises himself for a swoop. He comes
rushing down on quivering pinion. Nearer! nearer! It is a turkey buzzard,
who has scented a dead horse.
Constituencies can only be found where civilization rages.
CHAPTER XXXII.
ANOTHER CHANGE.
The world seemed coming to an end, I mean my world. I had “ran for
office” and was not elected, I had lectured and the people did not call for
more, my mines and all they contained were still under ground. The cities I
had planned were still unbuilt, I had written for our county paper and
gained a small county, but cashless reputation. The fall of 1866 was at hand,
and I was saying “Vanity of vanities, all is vanity,” when one day I received
an unexpected letter from the publisher of a San Francisco weekly paper
(The Golden Era). He said in substance, “Come to San Francisco and try
your chances on the Era. We will do the best we can for you.”
I went and was met by the good and great-hearted Joseph Lawrence, the
principal publisher, and up to that time an entire stranger to me.
The transformation in my life was sudden and startling. It was from the
mountain solitudes to the bustle of a great city, from the miner’s cabin to
the elegancies of the first-class hotel in which my friend positioned me;
from the society of the “boys” to that of artists, actors, editors, and writers,
some since of world-wide reputation.
It was the sharpest corner I had ever turned in my life. It led into a new
road, a new life, new associations, new scenes, and eventually new
countries.
And this change came sudden, unexpected at the “darkest hour” and like
“a thief in the night.”
San Francisco had changed greatly since I had left it eight years
previous. Much of the old “ ’49” characteristic had disappeared or was
disappearing. The roughness in garb and manner had abated, the high silk
hat topped more masculine heads, the afternoon feminine promenade on the
main shopping streets was more elegantly attired, “society” was classifying
itself into sets and “circles” more or less pretentious, many more men had
homes to rest in at night, the glare and splendor of the openly public