I was interested in the topic of banknote protection elements after reading an advertising article on geektimes that described a pocket gadget that recognizes the special element “And” (anti-stokes luminescence).

The world of methods for protecting documents and banknotes from counterfeiting turned out to be very interesting and exciting. In what, even now I understand Viktor Ivanovich Baranov.

As it turned out, information about specific elements and chip protection is very scarce. And those articles on the Internet that can be found, as a rule, even the illustration uses the same.
I didn’t "discover" anything fundamentally new, but it is possible that the results of my experiments will seem interesting to someone.
')
Conventionally, for myself, I divided the verification methods into two groups:

1. Protection elements available for testing without special tools (maximum magnifying glass).
2. Research using “instrumental” tools (UV, IR sources, etc.)

Topic on p1. it is simply huge and interesting, but it is relatively widely covered in open materials and I do not touch on it in this article.

For the experiments the instrument was assembled:

1. STM32F103ZET6 core board
2. LCD panel with a resistive touch screen.
3. OV7670 camera with FIFO AL422 buffer. The IR blocking filter is replaced by a filter that blocks visible light (a piece of non-exposed developed color film).
4. Magnetic head from magnetic card reader.
5. UV LED 3W 365nm
6. Infrared LEDs at 940nm and 850nm.

The choice is not the most optimal in terms of size, etc. But all this just lay idle, having remained from past projects.

The camera turns into an infrared near range by removing the IR filter and replacing it with a piece of developed color film (not exposed film tail).

Magnetic tags

To track the presence of magnetic tags, I used a head from a magnetic card reader and a two-stage amplifier on an LM358 op amp. Digitization of the signal and output to the screen as oscillograms to evaluate visually.

The gain factor was selected by an experimental method. The degree of magnetization of a card with HiCo stripe is much higher than that of a banknote.



What can be said from experiments with such a primitive sensor? Plots with magnetic properties in banknotes. It's all. To fake a bill for such a sensor, in my opinion, is elementary.

There are references (patents) about the following options for checking the authenticity of tags:

1. Through the degree of coercivity paint.
2. Magnetization retention time.
3. Hidden tags (information like a magnetic stripe card).

He experimented with magnetization with a permanent magnet. The results are not statistically significant, since the gap between the note and the head greatly influences the output signal. For this reason, I did not experiment with the writing / erasing head. Without normal uniform pulling and tight clamping, experiments give too much variation in values.

I saw photos from the documentation for professional instruments that show in 2D a color image of the degree of magnetization. The same documentation casually mentioned information that can be encoded in the magnetic labels of documents and which is visually visible in this device.

I tried the way that the adjustment of magnetic heads in embossers is often checked. Magnetic powder (toner from a laser printer comes down) on a magnetic strip.

The method allows to evaluate both the quality of the recording and the deviations in the location of the tracks, etc., etc.
Alas. The bill did not work. And rough paper / paint instead of a smooth surface and magnetization is not the same.

He could not determine if the "secret" encoding on the magnetic labels of bills. Most likely no.

Infrared image. Metamer paint

Perhaps the most common method of control in detectors. According to some information (examples of fakes) they learned how to fake in the early 2000s.

Euro, RUB

For Russian and banknotes and euros, no surprises and surprises. All exactly the samples on the site of the Central Bank. However, there are some tricks / recommendations:

1. Additional check - viewing on a gleam by IR radiation. It is mentioned that sometimes counterfeiters do not bother to fake the image obtained by "the light". And it differs in detail from the image in reflected light.
2. It is necessary to pay attention to the combination (in ordinary light) of pieces printed with metameric paints and ordinary ones. I did not check it myself. There are no counterfeit bills as a sample. But it seems to me that if someone bothered to get the metameric paints and make a fake, then with geometry it will not be deceived.
3. Clear edges in infrared light.

USD

The “old” USD 100 banknotes (1996 and 2006) caused some confusion. At first I even thought, “but not fakes”.

On one bill there were no “white” bands at all, on others there were “dirty” and with indistinct edges. It turned out, after digging on the Internet, it should be so.

The banknote, without any marks in infrared light - turned out to be a sample of 1996.

Perhaps due to the fact that there are no IR tags on these banknotes, this causes problems when exchanging these bills in many countries (tourists face this in the United Arab Emirates, China, Thailand, Maldives).

There is no fundamental difference between the banknotes of the series 2006 and 2009 (lilac) in the IR "color".
But the stripes on all new banknotes are much more contrast and with clear edges.

Special element "M"

The “M” special element is a part of an image printed with paint that is practically “white” in the light of the IR 940nm source and absorbing (“black”) in the light of the IR 850nm.

It is not found on all types of bills, and you need to know in what designs and denominations it should be, and in which not.

Label "M" is a very simplified method of spectral analysis of paint. A more correct way is the spectral analysis of inks used in professional equipment for examination:

1. In a closed box, shooting is performed in the rays of consecutively switched light sources (up to 15 pieces from UV = 200 to IR = 1200 nm) with a relatively narrow spectrum (LEDs, filters and lamps for UV).
2. A “color” image is built from images and is compared (algorithmically or “eyes”) with the standard. The method is not fast and the equipment is expensive.

Judging by some hints, the treasury stamp on US $ has a very distinctive “color” in such an analysis for different denominations of bills.

RUB

For RUB bills, the label is very contrasting and can be clearly seen if you alternately highlight 940nm and 850nm.

I did not find any information about fakes. But, for some reason, not all bills are used. On some designs there, on some not.

For example, it was not on the bills of face value 5000. Appeared on the bills of 5000 sample 2010.

On 500 ruble notes found at all. Moreover, this paint is printed number. And not a speck on the left side of the bill, as on most images found on the Internet.

For 1000 ruble two different designs did not find.

Euro

The "brightness" of a banknote in the light of 850nm and 940nm "by eye" is the same. It seems that this method of protection does not apply.

USD

There is no reliable information on USD banknotes. Therefore, I speak only for those banknotes that I had on my hands and which I checked myself.

Banknotes USD new sample 2009 (portrait without frame) $ 100, $ 50, $ 20, $ 10

Print ink color is dark green. The "brightness" of a banknote in the light of 850nm and 940nm "by eye" is the same.

USD $ 100 (large portrait in an oval frame) sample 2006 and 1996

The same color gif with the flashing seal of the treasury roams the Internet. In real life, the difference in brightness of the printing ink is not so noticeable, but it exists.

I suspect that this was not intended as a special feature, but simply a feature of the spectral characteristics of this paint. A feature that was just noticed on the Russian detectors with the analysis of the special element "M".

But, in principle, you can rely on this as a simplified spectral analysis of the paint in reflected light.

USD $ 1 and $ 2 banknotes of 2006 sample

The print and the number under it is very clearly tracked. Almost the same as on the Russian banknotes. Bright green print colors and numbers in white light.

Ultraviolet

Viewing in UV radiation is considered to be unreliable way (the first is fake). In addition, in daylight it is very difficult to see something. Although the way and simple, but "uncomfortable."

And, probably, a fake bill printed on paper that shines in UV and without UV can be seen immediately as a fake.

The LED (noname) is declared as 365nm rather closer to 400nm or simply has a strong illumination with violet visible light. From this poor contrast and quickly tired eyes.
For a professional LED detector, UV illumination would not be convenient. Although, perhaps, I came across such LEDs (from two different orders on eBay).

Euro and new USD

The picture as a whole corresponds to the declared reference images, but it should be noted that the image for the eye is not so contrasting as when illuminated with a UV lamp and the eyes quickly get tired.

To see the beautiful glowing stars when using LED UV with a strong violet illumination - you need to strain and turn on the silent imagination.

USD $ 100 new and old

Pink strip is better to look at the light. I don’t understand why it’s done so that on smaller bills the band glows much stronger than on large ones. This is inconvenient. The strip is not as visible as expected.

RUB of the nominal 1000 and 5000 of the sample 2010

Accidentally found that a piece of the image has the properties of phosphorescence after pumping UV. If you bring the banknote section close to the UV LED, and then turn off the LED / take the banknote to the side, then the banknote section glows with a bright green light.

Glow fades in 2..1.5 sec.

Anti-Stokes (Anti-Stokes luminescence)

Search for the Russian-speaking segment by the word "Anti-Stokes" gives only links to manufacturers and resellers of equipment or gadgets for checking banknotes.

With the indispensable inscription about the uniqueness of the technology, that “allows us to unambiguously judge the authenticity of the banknote”, that “there is a protective sign that could not be reproduced by anyone” and even that this technology is secret and strictly protected at the state level.

But the first thing that comes to mind: Paint with Anti-Stokes effect is used everywhere. Starting with bills of small denomination, ending with disposable excise labels.

Those. this cannot be an exclusive technology and there is where the counterfeit material is used for counterfeiting.

A search for "anti-stokes luminescence" gives much more information:

1. References to the sellers (Alibaba) of the “powder” of the phosphor for $ 700-1000 per kg. To choose from, with different re-emission spectrum. Most likely, the "man from the street" will not be sold, but the fact itself.
2. Scientific articles and patents on the Anti-Stocks effect.
3. Various "stuff" for sale, where the paint is used with anti-Stokes effect. As, for example, markers for detecting IR laser radiation. Very handy for setting the focus of the laser or setting the position of the “spot” of the IR laser. I even ran across samples of business cards with Anti-Stokes tags and powder for fingerprints.

Methods for detecting anti-stokes effect

In professional equipment for experts, this is most often referred to in the context of ruble banknotes. It seems no one except Russia and China (new yuan) does not use this method of protection.

Visual tracking of IR irradiance (from 900nm) laser

The IR laser module is relatively inexpensive and easy to use. The relative low price, if you buy a laser module, and not a ready "flashlight" on which sellers in Russia clocked a minimum of 400% of the profits.



The main disadvantages of the form factor "flashlight":

1. The 30mW laser glow is weak and difficult to distinguish in daylight (even in artificial light it can only be seen in the shade).
2. A small spot. Labels with Anti-Stokes effect, there are other places bills. But they are small and difficult to find and see in the laser beam without a magnifying glass.
3. The laser, even with a power of 30mW, is dangerous to the eyes in a direct hit. Especially when you consider that the pupil does not respond to IR.
4. It is not possible to distinguish visually tags of different chemical composition with similar spectra.

These shortcomings (except the last) are partially compensated by a magnifying glass with a laser and closed walls.

For example, very small luminous spots are scattered around this area of ​​bills 1000 RUB. If you do not know - you will not find. And you can only consider a magnifying glass and in the dark.

Photograph failed. But they are there. it may be a production defect. And maybe so conceived.



Professional equipment for experts has a powerful (non-laser) IR backlight (900-1100nm) for the manifestation of the effect on the entire surface under study.

Automatic detection of Anti-Stokes tags

In the descriptions of banknote counters and automatic detectors, the “pulsed mode” of the laser is mentioned (unfortunately, without details and without references to patents). Moreover, it is argued that the authenticity of the material of the label is controlled. Perhaps this is the principle of US Pat. No. 4,047,033

Option with a pair of IR LED and IR photodiode (or phototransistor?). Some sellers of such devices claim that the authenticity of the tag material is also controlled. On the pictures on both sides of the pocket detector board, the IR LED control circuit (transistor switch) is clearly visible and it looks very much like a phototransistor / photodiode in a darkened compound. (US 20040031931 A1)

After analyzing all the available information (patents), I deduced for myself the assumptions:

1. A reliable way to distinguish the Anti-Stokes tags of the “same glow color” but different chemical composition is a comparison of the re-emission spectrum. However, this requires high laser power and a high-quality spectrometer (the spot of light from a 50mW laser is not too bright).
2. Tracking anti-Stokes tags using an IR photodiode is most likely based on an “afterglow” effect (fluorescence with a lifetime of <100ns). Some patents mention 20..100ns for Anti-Stokes phosphors with lanthanides.

The presence of the effect of phosphorescence in IR was not too lazy to check and came to the conclusion that the STM32F103 ADC is not enough.



In principle, the synchronous detector circuit for cutting out a strictly defined signal band from a photodiode is not complicated. But in order to understand what to analyze, you need a good digital oscilloscope.

The test circuit is simple: turning on the IR LED for a short time (1ms) to the maximum current and analyzing (with a resolution of at least 1ns) the signal from the photodiode in the area immediately after turning off the LED. Unfortunately, I don’t have a good oscilloscope, and it’s a pity to spend about $ 1000-1500 to satisfy my curiosity.

So for me the question of the principle of detecting a tag with a pair of IR LED / photodiode remained open. As is the question of the credibility of claims that this method can be used to distinguish Anti-Stokes tags with different chemical compositions (the uniqueness of the raw materials used by the Goznak, etc.) and the same IR laser beam.