While 'fever detection' is new to most people, the technology has been practiced for many years and has a series of international standards that we explain herein.
In particular, the world's two top international standards groups, the IEC and ISO, have published 3 standards covering fever (i.e., febrile) screening:
IEC 80601-2-59:2017 Medical electrical equipment — Part 2-59: Particular requirements for the basic safety and essential performance of screening thermographs for human febrile temperature screening
ISO/TR 13154:2017 Medical electrical equipment — Deployment, implementation and operational guidelines for identifying febrile humans using a screening thermograph
ISO 80601-2-56:2017 Medical electrical equipment — Part 2-56: Particular requirements for basic safety and essential performance of clinical thermometers for body temperature measurement
We purchased those 3 standards for 530 Swiss Francs (~$550 USD), studied them and prepared this guide to explain how they apply to fever sceening, including:
Focus Area: Inner Eye, Other Areas "Unreliable"
ISO Face Examples Show Importance of Inner Eye
Eyeglasses, Hair Shouldn't Obstruct Face
Impact of Obstructions Shown by ISO
Manufacturers Contradict Standards
'Face Detection' AI and Temperature Algorithms
Minimum Lab Accuracy Level
Faces Must Be Parallel to Camera
Minimum Face Pixel Requirements
Secondary Measurement Required
Clinical Thermometer Requirements
Ambient Temperature Must Be Controlled, Ensure No One Is Sweating
Avoid Bright Lights, A/C Drafts, Other Heat/Cold Sources
Secondary Screening Area Requirements
Fever Cams For Use "Under Indoor Environmental Conditions"
Long List of Facilities Recommended
Ease-of-Extraction Emphasized
No Toilets Nearby
Backdrop Requirements: Non-Reflective, Avoid Dark Colors
[***] ********* ************ **** **** **** at *****-**-*****, *****-** **** *** *** entrances ** *********under ****** ************* ********** with the intention of separating febrile from afebrile individuals to help prevent the spread of communicable diseases
[***] **** ******** ******** ******* ********** for *** **********, ************** *** ********* of * ********* *********** ******** ** be **** *** ***-******** ******* *********** screening ** ***********under ****** ************* ********** to prevent the spread of infection.
This is a great article and very useful in this emerging field. While reading this I was simultaneously on a conference call where it was announced what one customer's standards will be for fever screening:
ICI cameras used at entrance points - if above 97.7 degrees Fahrenheit this classifies them for additional screening due to low accuracy on this unit. From there a handheld unit is used for verification and if above 100.4 degrees Farenheit they are disallowed access.
This seems like it would be a very wide net. Since most people being screened will be coming in from outside and the weather is warming up this seems like a serious concern with accuracy.
After the first 1,000 or so people processed, I would wager they change their approach on that, or give up on it altogether. They will be spending a lot of time dealing with false flags into secondary screening, though if the handheld unit is doing forehead temp at >100.4F, I doubt they will get many confirmations there, as people would have likely cooled down enough from standing around waiting in the long queue for secondary measurements.
While I agree these are sensible, I disagree no customers will follow them. We've taken multiple steps to ensure our sales team doesn't sell any system, unless the customer acknowledges and agrees to take a lot of these precautions (We did not mention bathrooms or lighting). Not in a fine print contract on page 300, but an up front discussion and then in writing on page 1 of the quote in clear bullet points saying "x,y,z" need to be done to get accurate readings. We also go to great lengths saying that if any of those points aren't followed to expect accuracy to plummet. I put the responsibility on the integrator to educate the customers on what is reality vs. marketing, also then following through these steps with the installation.
In this particular current case, we should bear in mind that current evidence suggests that the vast majority of people are asymptomatic, meaning of course that they exhibit no symptoms. So detection and possible exclusion of individuals presenting with fevers could, even in a perfect world with all standards adhered too, fail to secure a property and is for the purposes of "security theater" as much is anything else.
Was watching news last night and they had Spokane Airport on - they had installed Athena (not mentioned by name but could see their logo) and were scanning all passengers for temp at the TSA screening point. This was part of a larger story on all the airlines requiring masks for everyone. Story said they were first airport to implement the testing. Laughed because the video
I just had a customer ask about this, so I've been scouring for companies that are marketing as ISO / IEC compliant, while following the guidelines in this article
The only one I found that does the inner eye as recommended above is Sensia out of Madrid (and whom I've never heard of till about an hour ago)
At least they look like an actual purpose built product as opposed to a Thermal imaging camera trying to masquerade as a fever detection device
Spectron IR has been stating these exact guidelines the entire time and use a highly sophisticated medical IR camera. They are stating a uniform background be required, people be scanned one at a time, removal of sunglasses, eye glasses, hats, masks over the nose, etc. They really are ahead of the other "making a profit" companies that are in this field. www.spectronir.com
I've never used or set up a thermal cam. I would like a brief explanation on how a blackbody device improves accuracy. I understand it's for calibration and every marketing pic I've seen keeps it in the camera's FOV during operation. Is that necessary? Is there any physical connection between the camera and the blackbody? Are there grades (good, better, best) for these calibration devices? Should these be evaluated by IPVM? And how do I know it's accurate? Thanks.
Is there any physical connection between the camera and the blackbody?
No physical connection between the camera and the blackbody in the 3 we are testing currently (Dahua, Seek, Sunell). And from our review of other system's documentation, that's normal. You typically set the blackbody to a specific temperature and then input that temperature in the software of the thermal camera / measurement device.
Are there grades (good, better, best) for these calibration devices?
We don't know enough about blackbodies yet to understand all the differences. Some are obvious, though, e.g., the Seek is at fixed temperature, others support varying temperature ranges (35C to 50C), etc.
And how do I know it's accurate?
One way we are and have heard others do it is to test against hand held IR thermometers 'guns'. We have a handful of them and, for the short time we have been testing so far, the blackbodies hold their temperatures at the set mark.
John - have you considered making this article public so it can be spread on social media to educate companies buying thermal camera products under the guise of identifying temperature detection?
Thermography (measuring specific temperatures with a thermal camera) is a practical application of Planck's Law.
The LWIR waves that a thermal camera sees can be affected by various environmental conditions, including absolute humidity, temperature, etc. When this happens, the wavelengths hitting the camera are not exactly the same as those emitted by the actual object (extreme summary, but you get the idea).
By setting up a blackbody you are telling the thermal camera that it can rely on the values for the pixels covering the blackbody to be a perfect reference. It can then compute the difference between the value it is actually reading, which will be less than perfect due to the atmospheric conditions, and the value it expects. This allows the camera to determine an offset to use for other temperatures (wavelenghts) to get a more accurate reading.
This is similar in concept to setting the whitebalance of an optical camera. When the optical camera receives slightly distorted colors, you can define an offset so that it corrects the spectrum to give you a more accurate color representation. On a thermal camera we are correcting for temperature instead of color. You want the blackbody reference constantly in the scene, as the environmental variables effecting the readings are less likely to stay constant than say an optical camera under a fluorescent light.
For thermal camera applications that are not dependent on accurate temperatures, like detecting a person along a fenceline, all we really care about are relative representations of "hot" and "cold", so blackbodies do not normally come into play.
This use of a black body reference in the field of view is a good concept. However if the camera has drift concerns, then how do you know the field of view is uniform, especially where the black body reference is at? Because many of these companies are placing the black body so far away from the actual camera there is not many pixels covering the reference. If that area is non uniform because of drift, then what are you doing?
The black body is supposed to be at the same range you intend to measure at. Of course there are lots of ways to set it up wrong, and at some point you just can't compensate for poor design or installation.
Thermodynamic black bodies are not equivalent to optical black bodies that are pixel rendered, digitally digested and converted from radiation to resolution(false). Nice try BK, please keep your cameras lens out of physics. If a hardware only POC comes up we will seek you and your wikipedia wisdom. OH...case closed.
Agree? or Disagree.
Fatal googling has been known to one shot redditors...
#ArmednDangerous #Wolfhowl
No need for me to change bait on this one!
Simple, Max used math, not a lens'd thermal camera with what? ONVIF support LOL?
....Money for nothing, and your chicks for free. Oh Gee look at all the informative clickers with no clue on fact being facts?
This is article is very helpful. As an integrator, we must comply with this highly researched international standards. I deem that it is our moral obligation to educate the end users to be aware of the pros and cons of what product we recommend for this particular thermal camera application.
Thanks very much for the posting Charles and John.
I've been inundated by requests for clarification on the various screening options. I've also spent much time trying to correct some unfortunate misinformation that was spread by some manufacturers and integrators trying to take advantage of anxious end-users. The details you've provided regarding the standards will be of particular value.
Thanks for this very thorough article, really matches what we are trying to impose to our customers willing to deploy these scanning systems, with quite some difficulties I must say, as they really expect measures done on flows of poeple, thanks to the "good" marketing of the companies and vendors you mention.
I'm doubtful about several points however: ISO indicates that masks and face wash can impact the measure, and that's true in their field of analysis (measure done at upface level).
However, how can face washing impact the temperature of the inner canthus whatsoever ? I mean you're not washing the inside of your eye, and if some water drops get into it, the temperature goes back to normal after only a few seconds only (my guess though).
Also, it's true that wearing a mask causes your breathe to perturbate the measure, however as your breath is at your body temperature, couldn't that be instead an improvement ? I understand that the difference between the inner canthus temp and body temp is around 1 degree (celsius), so threshold for that measure is one degree lower than actual body temp expected, leading to more false positives if masks are used.
If anyone can explicit or correct me, I would be grateful !
The ISO and IEC don't explain exactly how cold water affects the inner eye. However, I would stress that the key zone of the face for body temp detection is technically "the region medially adjacent to the inner canthus of the eye", so it's not just the tiny fleshy bit of the inner eye but the skin around it as well, which would definitely cool down with some water.
Also, it's true that wearing a mask causes your breathe to perturbate the measure, however as your breath is at your body temperature, couldn't that be instead an improvement
As you noted, face skin temperature is not the same as body temperature - it's typically a few degrees lower, not just 1 degree (see OptoTherm's image below on "skin temperature offset"). So if someone's face reads 33°C, a hot breath exhalation will make it look warmer on the screen.
Hope this helps, any other questions or follow-ups let me know.
IPVM has done a great job of highlighting inaccurate and false claims regarding thermal screening, etc, and shown plenty of suspect installations. With that said, would it be possible to profile an accurate installation, to include the processes and contingencies in place?
Correct me if I am wrong. Per the IEC spec, I didn't see any fever detection thermal cameras on the market which can meet all the requirements.
Agree
Disagree
Informative
Unhelpful
Funny
Create New Topic
Login to read this IPVM report.
Why do I need to log in?
IPVM conducts reporting, tutorials and software funded by subscriber's payments enabling us to offer the most independent, accurate and in-depth information.
Comments (32)
Undisclosed Integrator #1
This is a great article and very useful in this emerging field. While reading this I was simultaneously on a conference call where it was announced what one customer's standards will be for fever screening:
ICI cameras used at entrance points - if above 97.7 degrees Fahrenheit this classifies them for additional screening due to low accuracy on this unit. From there a handheld unit is used for verification and if above 100.4 degrees Farenheit they are disallowed access.
This seems like it would be a very wide net. Since most people being screened will be coming in from outside and the weather is warming up this seems like a serious concern with accuracy.
Create New Topic
Undisclosed #2
You just wiped out about $500M of pipeline that a bunch of hacks were really counting on.
Create New Topic
Undisclosed Integrator #3
While all of these are sensible, no customer will follow them. It's going to be a lawless wasteland for the next 2-3 years.
Create New Topic
Undisclosed End User #4
In this particular current case, we should bear in mind that current evidence suggests that the vast majority of people are asymptomatic, meaning of course that they exhibit no symptoms. So detection and possible exclusion of individuals presenting with fevers could, even in a perfect world with all standards adhered too, fail to secure a property and is for the purposes of "security theater" as much is anything else.
Create New Topic
Undisclosed Manufacturer #5
Was watching news last night and they had Spokane Airport on - they had installed Athena (not mentioned by name but could see their logo) and were scanning all passengers for temp at the TSA screening point. This was part of a larger story on all the airlines requiring masks for everyone. Story said they were first airport to implement the testing. Laughed because the video
Looked for video - can't find it
Create New Topic
Undisclosed Manufacturer #5
Thanks U6 - That's the one
I just had a customer ask about this, so I've been scouring for companies that are marketing as ISO / IEC compliant, while following the guidelines in this article
The only one I found that does the inner eye as recommended above is Sensia out of Madrid (and whom I've never heard of till about an hour ago)
At least they look like an actual purpose built product as opposed to a Thermal imaging camera trying to masquerade as a fever detection device
Create New Topic
Undisclosed #9
I've never used or set up a thermal cam. I would like a brief explanation on how a blackbody device improves accuracy. I understand it's for calibration and every marketing pic I've seen keeps it in the camera's FOV during operation. Is that necessary? Is there any physical connection between the camera and the blackbody? Are there grades (good, better, best) for these calibration devices? Should these be evaluated by IPVM? And how do I know it's accurate? Thanks.
Create New Topic
Bobby Mancia Jr.
IPVMU Certified | 05/05/20 04:27am
This is article is very helpful. As an integrator, we must comply with this highly researched international standards. I deem that it is our moral obligation to educate the end users to be aware of the pros and cons of what product we recommend for this particular thermal camera application.
Create New Topic
David Sullivan
Really great article. Thank you John.
Create New Topic
Chris Clarke
Thanks very much for the posting Charles and John.
I've been inundated by requests for clarification on the various screening options. I've also spent much time trying to correct some unfortunate misinformation that was spread by some manufacturers and integrators trying to take advantage of anxious end-users. The details you've provided regarding the standards will be of particular value.
Appreciate the ongoing research!
Create New Topic
Undisclosed Manufacturer #12
This is probably one of the most well written articles on the subject ever.
Create New Topic
Undisclosed End User #13
Well done IPVM! Very helpful article.
Create New Topic
Laurent ASSOULY
Thanks for this very thorough article, really matches what we are trying to impose to our customers willing to deploy these scanning systems, with quite some difficulties I must say, as they really expect measures done on flows of poeple, thanks to the "good" marketing of the companies and vendors you mention.
I'm doubtful about several points however: ISO indicates that masks and face wash can impact the measure, and that's true in their field of analysis (measure done at upface level).
However, how can face washing impact the temperature of the inner canthus whatsoever ? I mean you're not washing the inside of your eye, and if some water drops get into it, the temperature goes back to normal after only a few seconds only (my guess though).
Also, it's true that wearing a mask causes your breathe to perturbate the measure, however as your breath is at your body temperature, couldn't that be instead an improvement ? I understand that the difference between the inner canthus temp and body temp is around 1 degree (celsius), so threshold for that measure is one degree lower than actual body temp expected, leading to more false positives if masks are used.
If anyone can explicit or correct me, I would be grateful !
Create New Topic
Undisclosed End User #14
IPVM has done a great job of highlighting inaccurate and false claims regarding thermal screening, etc, and shown plenty of suspect installations. With that said, would it be possible to profile an accurate installation, to include the processes and contingencies in place?
Create New Topic
Undisclosed #15
Please explain how the fever detection world.....
Protects us from a big beast like....
EBOLA?
yah? that's what I thought.
Might as well tell everyone with Plantar Fasciitus to move to the MOON. Oh hai Elon!
Create New Topic
Undisclosed Distributor #16
Great summary of IEC!
Correct me if I am wrong. Per the IEC spec, I didn't see any fever detection thermal cameras on the market which can meet all the requirements.
Create New Topic