Printing variability of copy detection patterns Roman Chaban Olga Taran Joakim Tutt Yury Belousov Brian Pulfer Taras Holotyak and Slava V oloshynovskiy Department of Computer Science University of Geneva Switzerland

2025-05-02 0 0 2.81MB 6 页 10玖币

侵权投诉

Printing variability of copy detection patterns

Roman Chaban, Olga Taran, Joakim Tutt, Yury Belousov, Brian Pulfer, Taras Holotyak and Slava Voloshynovskiy

Department of Computer Science, University of Geneva, Switzerland

{roman.chaban, olga.taran, joakim.tutt, yury.belousov, brian.pulfer, taras.holotyak, svolos}@unige.ch

Abstract—Copy detection pattern (CDP) is a novel solution

for products’ protection against counterfeiting, which gains its

popularity in recent years. CDP attracts the anti-counterfeiting

industry due to its numerous beneﬁts in comparison to alternative

protection techniques. Besides its attractiveness, there is an

essential gap in the fundamental analysis of CDP authentication

performance in large-scale industrial applications. It concerns

variability of CDP parameters under different production condi-

tions that include a type of printer, substrate, printing resolution,

etc. Since digital off-set printing represents great ﬂexibility in

terms of product personalized in comparison with traditional

off-set printing, it looks very interesting to address the above

concerns for digital off-set printers that are used by several

companies for the CDP protection of physical objects. In this

paper, we thoroughly investigate certain factors impacting CDP.

The experimental results obtained during our study reveal

some previously unknown results and raise new and even more

challenging questions. The results prove that it is a matter of

great importance to choose carefully the substrate or printer for

CDP production. This paper presents a new dataset produced

by two industrial HP Indigo printers. The similarity between

printed CDP and the digital templates, from which they have

been produced, is chosen as a simple measure in our study. We

found several particularities that might be of interest for large-

scale industrial applications.

Index Terms—Anti-counterfeiting, copy detection patterns,

digital offset printing, printing variability, substrate, similarity

measure, fakes.

I. INTRODUCTION

In recent years, copy detection patterns (CDP) [1] became

an attractive and popular technique for product protection

against counterfeiting. CDP are often used for the protection

of packaging and security labels. CDP are also used for

the protection of pharmaceutical products and vaccines, for

example, those against COVID-19 [2]. In general, CDP are

printed on digital off-set printers but classical off-set and ﬂexo

can be used as well. They are easily integrated in the package

design as shown in Fig. 1 according to [2], [3].

Besides the popularity of CDP and its wide usage, there

are still some issues that are little studied and raise questions

about CDP security in critical applications. This relates to

the lack of large-scale public datasets produced on industrial

equipment. Moreover, there are numerous researches which

aim at challenging CDP security by creating high-quality fakes

[4] [5] [6] [7]. Such fakes represent a considerable threat to

CDP as a protection technology.

S. Voloshynovskiy is a corresponding author.

This research was partially funded by the Swiss National Science Founda-

tion SNF No. 200021 182063.

Fig. 1: An example of CDP integration into a structure of QR

code as suggested by Scantrust [3]. CDP is placed inside the

QR-code and the copy detection is performed along with the

decoding of information stored in the QR-code.

In this paper, we aim to fulﬁlling the gap in the availability

of research datasets addressing the variability of CDP and

release a new public dataset of CDP printed on two indus-

trial printers HP Indigo 5500 (HPI55) and HP Indigo 7600

(HPI76) under various settings1. The main goal of this study

is to investigate the statistical variability of CDP in terms of

the deviation of statistics of printed CDP with their digital

counterparts in the accordance to the previous researches [8].

Each factor of variability is formulated as a separate research

question in this work. In summary, we investigated the impact

of the following factors:

•The difference in CDP statistics produced by two indus-

trial printers HPI55 and HPI76;

•The inﬂuence of substrate;

•The variability in continuous printing and over time;

•The effect caused by the acquisition device;

•The impact of deviation of printing resolution.

Taking into account the mentioned impact factors the main

contributions of the paper are twofold:

•The new public dataset with synchronized CDP and their

digital templates;

•The analysis of CDP variability of the above factors.

II. DATASET

In this work, we present a new public CDP dataset speciﬁ-

cally created to investigate the CDP variability. The produced

dataset of CDP contains digital templates of size 228x228

pixels with 50% of pixels being white and 50% black. The

digital CDP with the synchronization patterns are allocated on

A4 pages. One page contains 12 rows and 12 columns of CDP.

The resulting amount of CDP per page is 144. The basic set

1The dataset is available https://github.com/sip-group/snf-it-

dis/tree/master/datasets/indigo1x1variability.

arXiv:2210.05343v1 [cs.CR] 11 Oct 2022

Printing

resolution (dpi) Printer Printing

session

Printing substrate

IG AD AT CS

812.8

HPI55

0 1 / 1 - - -

1 10/4 1/1 1/1 1/1

2 1/1 1/1 1/1 1/1

HPI76

0 1 / 1 - - -

1 3/1 1/1 - 1/1

2 1/1 1/1 - 1/1

813 HPI55 1 3 / 1 - - -

HPI76 1 3 / 1 1 / 1 - 1 / 1

TABLE I: Produced data set: the numbers in cells with the

slash indicating ”p/s” denote pthe maximum number of print

runs and sthe maximum number of scan runs.

consists of 10 A4 pages with unique CDP codes. This set of

10 A4 pages is then printed with different printing parameters

and substrates at different time intervals on two printers. The

methodology of our approach for CDP production in details

is described in [4].

Taking into account that the methodology of CDP’s printing,

acquisition, and the following post-processing, i.e., synchro-

nization, grayscale conversion, normalization, etc., is the same

as in the previously released Indigo 1x1 base dataset [4], we

complement our new dataset by the subset of CDP from the

Indigo 1x1 base dataset. To show the importance of the CDP’s

variability for the authentication, we take into consideration

not only the originals but also the ML-based fake CDP from

Indigo 1x1 base dataset.

The subsets taken from the Indigo 1x1 base are marked

with printing session 0 and contain 720 unique CDP. The goal

of this complement is to investigate the printing variability

in different time moments while keeping the rest of the

parameters ﬁxed.

We investigate the following parameters of variability:

•Printers: HPI55 and HPI76.

•Printing sessions: 0, 1, 2. Session 0 was printed in May

2021, session 1 in November 2021, and session 2 in

February 2022.

•Printing substrates: Invercote G (IG), Algro Design (AD),

Atelier (AT), Conqueror Stonemarque (CS).

•Printing resolution: 812.8 and 813 dpi.

•Print run: 1 - 10.

•Scan run: 1 - 4.

We have chosen two printers to be coherent with our

previous studies. We have chosen the mentioned substrates due

to their popularity in the production of products’ packaging

as IG,AD, and AT substrates have roughly the same surface

structure, coating, and density. Moreover, we add the substrate

CS to the analysis as it has a rich texture and complex surface

structure.

The overview of the different sets in the produced dataset is

presented in Table I. Examples of printed and acquired CDP

reﬂecting the variability due to the impact of the substrate are

shown in Fig. 2, the impact of time printing sessions for the

Fig. 2: Examples of CDP printed on two different printers and

on three different substrates.

Fig. 3: Examples of CDP printed on two different printers on

IG substrate at three printing sessions.

Fig. 4: Examples of CDP printed on two different printers on

IG substrate during the same printing session. Three runs of

the same session are shown. One can notice quite essential

deviations between three runs.

same substrate in Fig. 3 and intra-session variability during

the continues printing in Fig. 4.

III. INVESTIGATION OF IMPACT FACTORS ON THE

VARIABILITY OF CDP STATISTICS

Instead of investigating the intra-class variability of CDP

statistics, we focus on an authentication setup, where the

variability of some similarity score between the printed CDP

and its digital reference used for the authentication decision is

studied. In this respect, the studied variability is a function of

the reference template and a chosen metric. Such an approach

has its advantages when directly evaluating the authentication

system performance in the space of chosen decision metrics.

Therefore, after the printing, acquisition, and pre-

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

PrintingvariabilityofcopydetectionpatternsRomanChaban,OlgaTaran,JoakimTutt,YuryBelousov,BrianPulfer,TarasHolotyakandSlavaVoloshynovskiyDepartmentofComputerScience,UniversityofGeneva,Switzerlandfroman.chaban,olga.taran,joakim.tutt,yury.belousov,brian.pulfer,taras.holotyak,svolosg@unige.chAbstractCop...

展开>> 收起<<

Printing variability of copy detection patterns Roman Chaban Olga Taran Joakim Tutt Yury Belousov Brian Pulfer Taras Holotyak and Slava V oloshynovskiy Department of Computer Science University of Geneva Switzerland.pdf

共6页,预览2页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Printing variability of copy detection patterns Roman Chaban Olga Taran Joakim Tutt Yury Belousov Brian Pulfer Taras Holotyak and Slava V oloshynovskiy Department of Computer Science University of Geneva Switzerland

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: