Nude scanner 3d

Naked Home Body Scanner | Naked Labs®

$1,395 Temporarily Out of Stock

Purchase Naked

  • Naked is now shipping
  • New orders ship in about two weeks
  • 30-day risk-free trial (less S&H)
  • Currently accepting U.S. orders only


Buy Now

New orders ship in two weeks. Buying multiple units?

Naked Features

3D Body Model

Explore your unique 3D model with rotate, pinch, and zoom.

Side-by-Side Comparison

See how your body has changed by comparing any two scans

Body Fat %, Lean Mass, Fat Mass

Set a baseline scan and track how your body composition is changing


Track 10 circumferences over time

Data Visualization

Rich graphs and the ability to compare many different metrics over time

4-Up View

See a snapshot of your body from all four sides

Watch the iJustine Unboxing

What's in the Box?

Box 1 (small)

Size: 18. 5"L x 18.5"W x 6"H
Weight: about 40 lbs.

  • Rechargeable rotating scale
  • Hard floor scale feet
  • Carpet scale feet
  • USB-C charging cable
  • Weighted cover for mirror

Box 2 (large)

Size: 69"L x 17"W x 6.5"H
Weight: about 35 lbs.

  • Body scanning mirror
  • Two stabilizing pins
  • Baseplate for mirror
  • Allen wrench
  • 4 ft grounded power cord

Top FAQ's

What is Naked?

Naked is the world’s first 3D body scanner for your home. It helps visualize your body’s progress so you can stay on track to meet your goals, whatever they are.

How does Naked scan me?

Naked uses Intel® RealSense™ Technology and a process called stereovision to create your body model. The sensors use infrared light in a manner similar to an XBox Kinnect or your TV remote control, and the processor in the mirror stitches together your body model before running proprietary algorithms to extract your measurements and calculate body fat percentage.

How does Naked calculate body fat percentage?

Naked has developed a proprietary body fat algorithm validated against DXA using a combinations of metrics extracted from your scans. We will continue to refine this algorithm and add more data and metrics to improve body fat % accuracy over time.

How does Naked compare to other body fat measurement methods?

Contrary to popular belief, body fat measurement is not, in fact, a measurement. Rather it's an estimation based on assumed constants and often derived from another body fat method. Naked uses a proprietary algorithm that relies on a variety of metrics extracted from your scan to calculate Traditional bioimpedance scales are typically +/-8% accurate to four compartmnet models and can be especially susceptible to factors like skin conductivity and hydration. DXA, typically considered the gold standard, uses x-Ray to measure bone density and body composition with +/- 1% accuracy to higher compartments. Volumetric methods (like hydrostatic weighing and Bodpod) use displaced water or air to measure your body volume, with approximately +/- 1. 5% percent accuracy to. Naked uses a proprietary algorithm that relies on a variety of metrics extracted from your scan to calculate body fat with +/- 2.5% accuracy compared to DXA.

How accurate are the 3D body models?

Naked has a depth map accuracy of a 1/10 of an inch.

Is there any way to get scanned now?

If you are interested in getting scanned now, please email us at [email protected]

How is Naked powered?

The mirror is designed to be connected to a 110v grounded outlet via a 3-prong grounded cable. The scale is powered by a lithium-ion battery and charged via a USB-C cable that is connected to the base of the mirror.

Read More FAQs


30-day money back guarantee

If you purchase Naked and decide it’s not for you, you have 30 days from when it arrives at your home to request a return and receive a full refund less shipping cost.

1-year manufacturer’s warranty

Naked will repair or replace any units with manufacturer’s defects. This does not include user-caused damage. If you do damage your Naked or require repairs outside your warranty, we can fix it for you, but you will be responsible for parts, labor, and shipping.


$100 shipping in the United States.

We cannot ship to P.O. boxes or military (APO/FPO/DPO) addresses. We recommend shipping directly to the location where you plan to use Naked, as the boxes are large and heavy.

International orders

We aren't accepting new international orders at this time, but you can sign up to be notified of any updates.

Metrics | Naked Labs®

In the study of body composition, your body is divided into four "compartments": bone, water, fat mass, and fat free mass (i.e. everything else). All body fat methodologies, including Naked, measure or infer two or more of these compartments.

How we calculate body fat

Naked uses your biological sex, height, and several measurements from your 3D body model to estimate your body fat percentage. Our formula is based on an equation derived from DXA results (the gold standard in body composition testing) developed by the U.S. Navy. Our research has shown Naked body fat estimates to be within about 1.5% of DXA results, on average.

Other body fat methodologies are subject to variance from hydration and oxygen volume. Bioelectrical impedance, like InBody, footpad or handheld devices, is hyper-sensitive to the amount of water contained in the body. Hydrostatic ("dunk tank") and air displacement are sensitive to the amount of air in the lungs or air pockets in the hair or clothing. Using measurements to derive body fat means that meals, water intake, recent workouts, and lung volume will not significantly impact your results.

Lean mass and fat mass

Naked uses your body fat percentage and weight to estimate your fat mass and lean mass. Fat mass is the raw weight of the subcutaneous and visceral fat in your body. Lean mass is the weight of everything else — primarily, muscle, bone, organs, connective tissue, and water. These numbers paint a fuller picture of your body composition than body fat percentage alone. Because body fat percentage is a relative metric that is scaled by your weight, your percentage may hold steady while you see fluctuations in the raw numbers for lean and fat mass.

Gender, age, and race

Naked's body composition estimates already take your biological sex into account. By collecing data on users' age, race, and activities, we're working to improve the accuracy of our methodology for all different populations. Our hope is that this research will impact not only Naked's product, but also contribute to the broader field of knowledge.

Currently, we're working with our academic partners at The University of Hawaii, Pennington Biomedical Research Center, and University of Colorado to refine Naked's body composition metrics. This should yield improvements to accuracy — and new features, like regional body fat percentage — in the near future.

Read more

Dr. Sam Winter, Naked's in-house scientist, explains all the major body fat methodologies in her Body Fat Series on our blog. Check it out!

3D Body Scan DXA Water Displacement Air Displacement Bioelectrical Impedance Calipers
e.g. Naked Dual-energy X-ray absorptiometry "Dunk Tank" e.g. BodPod e.g. BIA, InBody, FitBit Aria "Skinfold test"
How Circumferences from the 3D scan are used to calculate DXA-based body fat Two X-rays of the body are used to visualize tissue distribution based on energy absorption rates The volume of water (at a given temp) displaced by the body is used to calculate fat mass The volume of air (at a given temp) displaced by the body is used to calculate fat mass An electric pulse sent through the body is used to infer fat mass, which is the most conductive tissue The thickness of the skin pinched at 3 to 11 places on the body is used to calculate overall fat %
Compartments 3 (inferred) 3: bone, fat mass, fat free mass 2: fat mass, fat free mass 2: fat mass, fat free mass 1: (sort of) total body water 1: fat mass
Pros Includes 3D scan
Visual feedback
Private, at-home
Takes 5 mins
No help required
“Gold standard”
Regional fat and lean mass
Visual feedback
Can stay clothed
Former “gold standard”
Large population dataset
Large population dataset
Easier than dunk
Can stay dry
Fast results
Available on consumer-grade smart scales
Can stay clothed
Readily accessible
Can stay clothed
Cons Sensitive to variation in clothing, hair, lighting and pose Requires travel
Not widespread
Mild radiation exposure
Overestimates athletic people
Requires travel
Less widespread
Wear swimsuit and get wet
Sensitive to air in lungs
Inaccurate for many races
Requires travel
Less widespread
Wear swimsuit and swimcap
Sensitive to air in lungs, hair and clothing
Sensitive to hydration, food and exercise
Does not directly measure any body tissues
Derived from 2-comp methods
Requires practitioner
Physically uncomfortable
Does not directly measure tissues
Derived from 2-comp methods
Accuracy* 2. 50% - 1-2% 2-5% 8-10% 6-10%
Cost $9/scan (scan weekly for 3 years on Naked) $45-$150/test + travel $30-$60/test + travel $40-$60/test + travel $20-$250 for smart scales $10-$50 per set
Read more about 3D scans Read more about DXA Read more about volume-based methods Read more about volume-based methods Read more about bioimpedance Read more about calipers
*Accuracy numbers are based on a subset of peer-reviewed journal articles using DXA as the criterion method.

Measurements are a useful way of tracking the body's changes beyond the scale by providing information on where fat loss and muscle gains are distributed. Naked provides circumferences at nine different locations.

How we measure

Once a body model is generated, Naked uses feature recognition algorithms to identify key landmarks on the body. Measurements are then drawn around a given body part, perpedicular to the primary bone. For each of the nine locations, Naked extracts a concave circumference, which follows any curvature or indentation on the body — like your spine, for example. This approach produces slightly larger circumferences than what you'd see with a tape measure.

Following your first scan, you will have the opportunity to adjust the precise measurement locations on your body model. This allows you to optimize the measurement site for your unique body and goals. Naked then tracks measurements at those exact same locations on every scan.

This feature-based anatomical approach makes Naked more reliable than other methods. Tape measures are difficult to place consistently and with the same tautness every time. Other 3D-scanners use the widest or narrowest circumference within a given bodypart — which can shift over time. Using either of these techniques, it becomes increasingly difficult to measure the same position on your body as you experience changes in fat and lean mass distribution.


Compared to mannequins and other static objects, Naked's measurements are accurate within 1.5 centimeters.

Providing a figure for accuracy of human measurements, however, is incredibly challenging given that all existing methods to measure circumferences are subject to measurement error. Even the most trained tailors will not be able to achieve a perfectly reproducible circumference, measured at exactly the same location with identical tautness of the tape measure.


Because Naked's measurements are generated along the surface of the 3D body model, the virtual 'tape measure' that generates Naked's measurements is extremently consistent. That said, Naked's measurements are sensitive to slight changes in pose, posture, clothing and even breathing — for scans taken back-to-back, normal variability is in the ~1 centimeter range.

Crackers learned how to make lockpicks using 3D printers




Lock picking experts have created special software that can generate a lock pick for almost any type of lock and print it on a 3D printer.

Mechanical lock picking experts Jos Wayers and Christian Holler have been able to develop special software that can automatically generate a virtual image of a master key for almost any type of lock. Moreover, this master key can later be printed on a 3D printer and open this lock with it.

All it takes to crack is a photograph of the keyhole, knowledge of the length of the lock cylinder, and the approximate location of the furrows inside the lock. Moreover, the second and third points are now easy to find on the Internet, Weyers and Holler pointed out. All this data is processed by a special Photobump program, which models the key.

The principle of operation of such a master key is very simple and has long been known to burglars: after it is inserted into the lock, it must be hit several times with a hammer to knock out the pins (blocking elements of the lock) into their channels and turn the key.

Researchers have clearly demonstrated this principle in the video:

What is new in this case is the simplicity with which you can create the right master key for the right lock. Just a photo from a smartphone and a couple of numbers. Fortunately, the creators of the software are in no hurry to share it, because they understand what potential harm it can cause. After all, it allows you to create master keys even to those locks to which, in principle, there are no keys.

Instead, Christian Holler plans to speak personally at the Lockcon trade conference to talk about his method.

Researchers are currently working with the German police to see if such keys can even leave clues.

However, even this is not the first way to print a 3D copy of a key: the KeyMe service has been operating in the USA for about a year now, which automatically recognizes the key parameters from a photo and creates a virtual copy that can be printed at special kiosks or ordered by mail. In the same way, the key parameters can be provided to the locksmith, and he will be able to restore it from them.

Zeus will be taught how to print in 3D

A device called Zeus has been developed in the USA, which combines a 3D printer, scanner and fax machine. Interviewed...

August 22 19:26

KeyMe is certainly not intended for burglars; on the contrary, it was created so that the owner of the house could quickly make a duplicate of the key in case of something (for example, if the original was lost). Actually, without the original key, it is useless: in order for the service to create a virtual key, you need to take a picture of the real one on a flat white surface (for example, on a piece of paper).

Using 3D printing technology, you can create a complete set of criminals, not only lock picks, but even firearms.

Thus, the American gunsmith-enthusiast Cody Wilson was able to print a pistol of his own design called the Liberator on the printer. Metal in it was only a striker made of a nail. All other parts are plastic from the printer. In the first two days, the Liberator model was downloaded over 100,000 times.

The pistol consists of 16 elements in total. Liberator is chambered for 9x17 mm. The creator's video demonstrates that plastic does not burst from a shot and that it can really be fired from:

find. But in any case, the printing of such weapons in many countries, including Russia, is illegal. Therefore, there is no information about other similar developments. The creator of Liberator Cody Wilson himself could well have been sued for distributing weapons.

Be that as it may, progress really does not stand still, and new technologies penetrate absolutely all spheres of our life, including criminal ones. And in everyday life, the danger from such inventions can be greater than the actions of hacker groups.

According to Anton Fedoseev, manager of educational programs at Autodest, to Gazeta.Ru, the legislative base and regulation in this area are just beginning to take shape, and at the same time, 3D printing is a rapidly growing and developing technology. “In the next 3-5 years, 3D printing and 3D modeling at the amateur level will become as easy and affordable as digital photography and printing at home.” From a technical point of view, nothing prevents us from making such a model of the same key and printing it in one of the prototyping centers, the expert specified.

Subscribe to Gazeta.Ru in News, Zen and Telegram.
To report a bug, select the text and press Ctrl+Enter




Picture of the day

Russian military operation in Ukraine. Day 239

Online broadcast of the Russian military special operation in Ukraine — Day 239

"Bypassed sanctions, laundered money. " In the USA they want to put the son of the governor of Uss for 30 years0005

Krasnoyarsk Governor Uss stated that the US accusations against his son have a "political color"

“I’m idle and I’m not going to quit”: how they work in Western brand stores in Russia

Former Zara sales assistant said that after his dismissal he was paid five salaries

FT learned about the shock in Kyiv due to the Republican warning about the reduction of support for Ukraine

Scholz: Germany freed itself from dependence on Russian gas

Putin signed a decree on the abolition of Rostourism and the transfer of its functions to the Ministry of Economic Development

RBC: Deputy Minister of Transport Sukhanov will resign after the explosion on the Crimean bridge

News and materials

Asos online store reduces costs due to falling demand for clothing from customers

Stoltenberg: almost all NATO countries have agreed on the accession of Sweden and Finland to the alliance

The ex-coach of Spartak Alenichev told which clubs of the First League he is ready to lead

Mishustin: anti-Russian sanctions of unfriendly countries hit them themselves

42-year-old soloist of the "Brilliant" Ksenia Novikova showed a figure in a swimsuit

A mural with jazz stars Butman, Lundstrem and Utyosov appeared in Moscow

Taylor Swift's ex-boyfriend joined the Ukrainian army

An 82-year-old pensioner with dementia from the Leningrad region, who disappeared for three days, was found alive

Mishustin called the intention to exclude the Russian Federation from global trade absurd

The release of the series with Pavel Priluchny “Union of Salvation. Time of Anger" suspended

A resident of St. Petersburg beat a friend who offended her to death with a frying pan and went to bed

In the Kursk region, 50% formed the composition of the mobile reserve for territorial defense

There are details of the theft of a police car in Novosibirsk

Volkanovski names two of the greatest Russian MMA fighters in history

Eurobond holders will ask the State Duma to oblige Russian companies to issue replacement bonds

A man burned cars out of revenge to his stepfather in Komsomolsk-on-Amur

58-year-old top model Elle MacPherson posed in swimsuit

Elon Musk wants to make Tesla more expensive than Apple

All news

How many Russians and Americans there will be by the end of the century: the forecast of the director of the Institute of the Russian Academy of Sciences

The head of the CEMI RAS named the only scenario in which the population of the Russian Federation will grow by the end of the century

Interview with director Evgeny Grigoriev about Deadline, ageism and Russian streaming

Director Evgeny Grigoriev spoke about working with Yura Borisov and Pavel Derevyanko

“I came home, and there was my mistress”: 5 famous women who caught their men cheating

Yulianna Karaulova and four more celebrities who revealed the betrayal of a partner

On the verge of a nervous breakdown: how to stop arguing with a child

Psychologist Surkova: the technique of free writing will help to cope with irritation with a child

"Diplomatic conflict resolution". Erdogan persuades Zelensky to negotiate

Erdogan called for the settlement of the conflict in Ukraine through negotiations

“The guys are raiding. That's all": fighters with "barrelers" were prosecuted for a fight

An "anti-barrel" activist spoke about a brawl with a traffic offender, because of which a case was opened

"Beautiful, bad, evil." Zac Efron - 35

"Fathers with many children were recalled from military units and returned to their families"

Volodin: fathers with three or more children were exempted from partial mobilization

Putin introduced martial law in the DPR, LPR, Zaporozhye and Kherson regions

The Federation Council approved Putin's decree on the introduction of martial law in four regions of the Russian Federation

"Corsage" - a hooligan drama about Elisabeth of Bavaria, which won the acting award at Cannes

Review of the Cannes hit "Corsage"

"For a number of operational considerations." Kyiv will not receive Israeli weapons

Israel refused to supply weapons to Ukraine

"The city turns into a fortress." Russia is preparing for the battle for Kherson

Stremousov said that the Russian military repelled all attacks of the Armed Forces of Ukraine in the Kherson region

"The goal is to capture the Zaporozhye nuclear power plant." What is known about new strikes on Energodar

Development of personal screening systems. Inspection equipment. Master Class. SPETSINTEK LLC

How to improve air traffic safety without disturbing passengers with a large number of checks?
The innovative security scanner QPS100/QPS200 will help you with this.

The typical air traveler has mixed feelings about safety. According to a survey presented by the German Federal Association for Information Technology, Telecommunications and Media (Bitkom), the majority of passengers approve of the widespread use of airport security technologies, including scanners. At the same time, they are annoyed by many inconveniences, ranging from luggage restrictions and the need to remove various items of clothing during security checks, to the unnatural position required by conventional body scanners and being searched by security personnel. And this is only exacerbated by the long wait at the checkpoints. With security standards unlikely to be relaxed any time soon, technology faces a new challenge to find a solution to this problem despite even tighter security measures.

Figure 1. The dual-panel QPS200 is designed primarily for airport use. It can be integrated into almost any security checkpoint. The QPS100 model consists of one flat panel, therefore, for a full inspection, a person is sanitized twice - in front and behind.

The goal is to maintain the highest level of safety while maintaining the greatest possible comfort for passengers. The new body scanner R&S QPS200 is an important step in this direction.

New product launched

The first scanners were developed in the early 1990s, before there was a significant market for them. These X-ray backscattering devices were used at airports. Microwave devices followed a few years later, although demand remained limited. After the dramatic events of September 11, it became clear that the market for security equipment would develop rapidly. As a result, Rohde & Schwarz very quickly decided in 2007 to become a partner in a consortium working on a project to offer a European alternative to products from the United States that were pioneers in this category. A three-year proof-of-concept research and development phase has concluded, which has validated the consortium's path. However, the divergent concepts of the serial production model have forced the participating partners in the industry to develop their own, separate paths. Within QPASS, a project initiated by Rohde & Schwarz and funded by the German Federal Ministry of Education and Research followed. Rohde & Schwarz, together with the Institute for Microwaves and Photonics at the University of Erlangen-Nuremberg, Germany, has been working to prepare for the production of a prototype that has shown the potential to meet all expectations, even as new EU requirements pose additional challenges.

In the past, devices were designed to create images that could then be evaluated by screeners. The fact that intimate details were also visible sparked an emotionally charged public debate that forced US and European authorities to change their rules. It is now forbidden to display photo images at checkpoints. Apparently, this is a simple regulation, but with extremely serious technical consequences. Now the task was to develop a method for automatically detecting suspicious objects. This required extending the millimeter wave technology with a fast imaging system that could reliably filter out anomalies from the measured data for display (Fig. 2). The development team assembled for this challenge made such rapid progress that the QPS100 production model was ready for deployment in 2014. Field tested, continuously improved, with all applicable certifications and expanded to include the QPS200, the R&S QPS platform today serves as an innovative security solution not only for airports, but for all areas requiring a high level of security and limited access. Demand is high. Countries such as Germany are following the example of the United States, which in 2010 began to consistently equip all airports with scanners. With the signing in July 2016 of the Agreement between the German Federal Ministry of the Interior and Rohde & Schwarz for the supply of 300 units, the R&S QPS200 is well on its way to becoming the standard equipment in German airports.

Figure 2. The scan result is displayed on a neutral graph. Personal data is not displayed or stored.

Brief history of airport security audits

The history of airport security, at least as far as defense against attacks, is linked to a series of key events and responses to those events. In the early decades of commercial aviation, airport security was essentially non-existent, as flying was an exclusive adventure that only wealthy clients could afford, and far from the stage of violent crime requiring protective measures. Even when there was air piracy, mostly politically motivated, with a record 82 hijackings in 1969, the authorities still did not see the need for significant countermeasures. Metal detectors were only used when examining passengers when there were grounds for suspicion. It wasn't until the 1980s, when the US in particular intensified its war on drugs, that passenger screening was stepped up and service dogs were brought into play. In 1988, the Lockerbie bombings resulted in some baggage being scanned in Europe; However, it was not until 2003 that checking every piece of baggage became mandatory under EU regulations.

In many ways, the attacks of September 11, 2001 represented "time to go" in terms of security policy. Every possible level of government has done everything possible to prevent re-attacks, from extensive cross-border reconciliation of passenger data to air marshals, to a ban on the transport of sharp or sharp objects such as nail files in hand luggage (The installation of such equipment cost only for Lufthansa alone, more than 30 million euros according to company statements). As a result, from this point on, there have been dramatic changes in the security check before boarding. Richard Reid's attempt in December 2001 to detonate explosives hidden in the heel of his shoes was thwarted, but the US and several other countries responded by requiring passengers to remove their shoes for inspection. A ban on large volumes of liquids was further introduced in 2006 after a plan to use liquid chemicals in an attack was revealed in the UK. During Christmas 20091999, terrorist creativity even extended to underwear when a Nigerian citizen tried to detonate explosives hidden in his underwear shortly before landing in Detroit. The Transportation Security Administration, set up in the US in response to 9/11, responded to the attempt by introducing full-body scanners at airports across the country. The installation of these devices, which were developed in the 1990s, began in 2007 at airports in the United States and other countries and cities such as Amsterdam, but their widespread use began in 2010. This first generation of scanners was based on X-ray technology, which made visible not only potentially dangerous objects, but also private parts of the body, leading to significant public criticism. In addition, X-rays have an ionizing effect and therefore cause damage to cells, even though the authorities rule out any health hazard due to the low level of radiation from the devices (measurements show that the level experienced by passengers is approximately equal to a few minutes of cosmic radiation). However, these backscatter devices have largely disappeared from the market and are now banned in many countries. Even the Transportation Security Administration took the big decision to eliminate all first-generation devices between fall 2012 and May 2013, replacing them with alternative millimeter wave technology. However, this move was not made for public health reasons, but rather because the device manufacturer failed to implement a software update in time, in which the scan results would be impersonal. In response to public outcry, US authorities included a provision in the Modernization and Reformation Act of 2012 that nude images created by body scanners should be replaced with a symbolic image of the body, identical for all subjects.

For health reasons, the European Union banned the use of X-ray scanners (and all technologies that use ionizing radiation) in a regulation that came into force in November 2011. The retention of personal data and data protection rights is also mandatory. Germany conducted the first test of the first generation mmWave scanner in 2010 at Hamburg Airport. Field trials were carried out and scientifically supported by the center of the German Federal Police. 800,000 volunteers, including the German Federal Interior Minister de Maizière, gave permission for the scan. The purpose of the large-scale test was to determine how the devices would perform in real life, to identify any problem areas and to figure out how to optimally implement common testing methods. National authorities work directly with the European Civil Aviation Conference (ECAC), which is responsible for air transport-related certification, and the Hamburg test has played a pioneering role here. The test results (including a high false alarm rate) have led the German federal government to conclude that the devices currently available on the market are not yet suitable for general use. Since then, manufacturers have stepped up efforts to improve reliability.

Latest generation scanners such as Rohde & Schwarz's QPS are smarter than their predecessors and suitable for unlimited and widespread use. Regulators also require scanners to speed up security checks rather than slow them down. The fast devices currently available make this possible. The increased use of automated safety equipment gives reasonable hope that the perceived burden on passengers will be reduced to an acceptable degree in the near future. The first step in this direction was the complete removal in the EU of the ban on large volumes of liquids in hand luggage in 2014 (although restrictions on medicines, special products and goods purchased in the duty free zone remained). This was made possible thanks to new technology that allows the detection of liquid explosives. Advances in scanning technology will soon allow screening through thick layers of clothing so that passengers no longer need to take off their coats and jackets. Solutions invisible to the eye are already on the horizon, meaning passengers will be subject to security checks without even noticing it. It is hoped that travelers will be able to board the plane without issue, unlike the woman who had her Christmas cake confiscated in Las Vegas because a security guard thought the icing looked too much like explosives.

New approaches

The QPS Scanner differs from competing products in the way it looks, how it works and how it scans. Vast spaces with bulky stands are a thing of the past. Visually appealing, compact, flat panels can be integrated barrier-free into a checkpoint (Figure 1). The open design gives security personnel an unobstructed view of the entire checkpoint. The scanning procedure, which is considered unpleasant due to traditional equipment, since passengers must keep their hands in the air, as if they were stopped by the police, is now much more comfortable for passengers. You need to slightly spread your arms in a natural position, which is possible even for physically weakened people, and is considered ethical in different cultures.

The QPS Scanner is a completely electronic, low noise instrument with no moving parts. It is the only device on the market that is based on the multi-static principle familiar from radar technology, where the reflected transmitted signal is applied simultaneously to a large number of receiving antennas (Figure 3). This provides better illumination of the scanned person, resulting in improved detection quality.

Figure 3. Multi-static principle of operation: While only one of the 3008 transmitters is active in the panel, all 3008 receiving antennas receive the reflected signal. Since each transmitter also quickly passes through 128 frequencies before the next one starts transmitting, each scan cycle transmits over a billion (!) complex values ​​to the panel (amplitude and phase), in less than 32 ms.

Apart from a calibration check every six months, the system is virtually maintenance free. Of course, these checks are carried out on the spot. Installation and commissioning take less than an hour, because all the preliminary labor-intensive processes are carried out at the factory.

Conventional microwave scanners illuminate objects at frequencies below 30 GHz. The R&S QPS, in contrast, achieves higher spatial resolution as it operates in the millimeter wave band between 70 GHz and 80 GHz, a band also used by vehicle parking sensors. At about 1 mW, the peak transmit power is about three times lower than that of mobile phone emissions and is hardly detectable where the person being scanned is standing.

Scanning volume - 1.9 mm × 1.9 mm × 5.7 mm voxels, which is achieved not only due to the high frequency range, but also due to the multistatic principle of operation. It was difficult to implement this in a high resolution, close range scanner. To make a multistatic mmWave scanner workable, its surface must be completely covered by antennas, which means placing a quarter of a million antennas on a panel for an R&S QPS size scanner. It took years of fundamental research in an attempt to reduce the number of antennas within reasonable limits, and this became the basis for more than one doctoral dissertation. As a result, a design was developed, consisting of 3008 transmitters and 3008 receivers, distributed in 32 clusters on each panel in a checkerboard pattern (Fig. 4). In addition to transmitters and receivers, the clusters contain the electronics required for frequency processing and conversion. The received signals are downconverted to 25 MHz and the signals from the four clusters are received to one of the eight IF modules located behind them. In IF modules, they are subjected to analog-to-digital conversion and processed into the original data in the form of images. At the last stage, the data is sent through two centralized boards and a PC adapter to the integrated inter-process communication (IPC) module, where the final analysis of the data takes place. Potentially dangerous objects are automatically detected and calculated, their exact location is displayed. Each panel also contains a synthesis generator for signal initiation, as well as auxiliary elements such as a battery, signal distribution, interface ports, and display elements.

Panel structure R&S QPS

Figure 4. The scanner has a modular design. The external RF interface is distributed in 32 clusters in a checkerboard pattern.

The QPS200 is designed for full scans in as little as 32ms per panel. Computer analysis currently takes a few seconds, and this time will be significantly reduced in the future through the use of more powerful data processing technologies. High-performance graphics processing units (GPUs), optimized by the graphics card industry for massively parallel processing, open up new possibilities on the road to real-time analysis, a prerequisite for future development.

Automatic detection task

At first glance, the body scanner works just like a classic photo booth. The person is illuminated, the reflected light is reproduced on the media (in the case of a scanner, data instead of paper) and an image is developed using the scanner software. However, a fully automatic scanner is not designed to create photographs for human viewing. Instead, its purpose is to provide an interpretation of the image, its evaluation, which is a much more difficult task.

The goal is both simple and complex: to detect all potentially dangerous objects that are carried on the body or in clothing, regardless of the type, size, location and composition of the material. procedures, that is, finding specific objects by pattern matching, will definitely fail. For example, there are too many materials and shapes that can be used as firearms or knives. The scanner had to be able to recognize specific objects based on their characteristics and purpose, regardless of shape, material, and this was a hopeless attempt. Therefore, the task of scanning is not to identify objects, but to identify areas on the body that, when assessed by analysis software, are significantly different from areas that do not arouse suspicion. The task of the developers was to teach the software to determine what is considered not suspicious. However, any attempt to do so on the basis of a "whitelist" of harmless things is also a futile task for the same reason as outlined above. Too many possible variations arise even with a simple example, such as menswear. Whether the person being examined is wearing a pullover or shirt, whether it is made of wool or synthetics, whether it closes with a zipper or buttons, which, in turn, can be large or small, made of plastic, metal or mother-of-pearl, they are located on middle or side, whether a lanky guy is standing in front of the scanner, or a hefty six-and-a-half-foot tall man, the scanner should evaluate all this and more without error and let them go through the examination procedure as usual. It was clear that a completely new approach would be required to solve this problem, which would combine the methodologies of image processing, machine learning and, very importantly, deep learning, a modern method of creating artificial intelligence.

Each scan creates a huge amplitude and phase dataset from the 3D scan volume that can be used as a basis for analysis. The scanner works similarly to a vector network analyzer. The received signals are compared with the transmitted signals in amplitude and phase, and the differential value contains all the necessary information about the test object. The challenge is to interpret this information through proper modeling, that is, by mapping the physical data into specific characteristics and attributes of the object. Examples of these attributes include signal intensity, surface roughness, and multiple reflection strength. In fact, over a thousand different attributes can be defined and combined into a high-dimensional attribute space. Suspicious objects or materials leave behind characteristic identification data in the place where they appear in certain subsets of attributes (=combinations of attributes) in certain ways. Classifiers are used to model these subsets (Figure 5).

Comparison of measured data with the characteristics of the object using classifiers

object. In this example, there are three classifiers, each with two attributes. Combining all classifiers gives the final result of the discovery.

Figure 6 demonstrates this principle based on a classifier that uses attribute signal intensity and surface roughness. In this two-dimensional attribute space, an unsafe area of ​​the body can be distinguished from an area in which, for example, a black powdered substitute was hidden under clothing. When deciding what exactly to highlight as a suspicious object, clear boundaries are an important criterion for the convenience of using the classifier.

Classifier principle

Figure 6. An example of a classifier with two attributes to determine the characteristics of a particular object. Attributes should be chosen so that a clear decision boundary can be defined; Two possible boundaries are shown here.

At the time the development of the QPS100 was started, it was still necessary to manually define and parameterize suitable attributes and classifiers (feature engineering). However, over the past few years, there have been huge advances in the field of machine learning, in particular deep learning. The latest version of the detection software has also been trained using deep learning algorithms.

Deep learning has replaced traditional machine learning algorithms in many applications. Google Android voice recognition, Facebook face recognition and Skype voice translation are based on deep learning. In early 2016, Google's AlphaGo computer program caught everyone's attention by beating numerous world-class Go players in a feat previously thought impossible. The game power of the software lay in the algorithms and analysis rules generated by neural networks using deep learning.

You can achieve amazing results with neural networks when it comes to pattern recognition. For example, networks especially suited for image processing, called convolutional neural networks, are now outperforming even humans at tasks like traffic sign recognition. In order for a neural network to complete a task, its topology and neuron switching functions must be perfectly tuned. Therefore, it took a lot of effort to develop the networks that underlie the R&S QPS software, and it also entailed huge costs. To find out if the network will work as intended, it must first be sufficiently prepared using data that meets high quality requirements**. In this case, the software was provided with a large set of labeled training data obtained by scanning many thousands of test subjects***. Labels are used to indicate any problem areas, if any, on the test object. The type of problem (knife, explosives, etc.) doesn't matter. From the vast number of samples processed, the software learns on its own what the alarming samples look like. In an extremely intensive processor optimization process that changes millions of parameters, the deep learning algorithm works through the database, finding those attributes and classifiers that are best suited to identify cases of violations. The solution to the problem obtained in this way is implemented in the scanner firmware. The overall learning and application cycle is highly unbalanced in terms of the time required. The extended learning phase (long calculations are performed on the GPU cluster) creates a program that, in actual use, can make decisions in seconds.

Deep learning provides decisive advantages in scanner software development. This method not only achieves high detection quality, but is also particularly attractive because the results are generated automatically. Software becomes a partner in development, freeing engineers from monotonous routine tasks. Instead, engineers can invest their time in improving neural networks and deep learning algorithms, leading to even better detection results.

Summary and future developments

With the QPS security scanner series, Rohde & Schwarz is opening a new chapter both for this category of devices and for aviation safety. Innovative hardware and software solutions meet the requirements of airport security workers while meeting the expectations of passengers to the maximum. Operators can integrate compact devices into a barrier-free checkpoint to ensure high passenger throughput.

Learn more