Dahua’s DH-TPC-BF3221-T thermal hybrid bullet camera is at the heart of the company’s BTM solution. It features 256 x 192-pixel VOx uncooled thermal sensor technology, with a high thermal sensitivity of <50mK, a focus free athermalized lens with a focal length of 7mm and an operating distance of 3 metres.

THIS DH-TPC-BF3221-T thermal hybrid bullet camera from Dahua has an interesting specification in its own right but it’s the camera’s ability to support Dahua’s body temperature management solution that’s most interesting to the market right now and is the subject of our demonstration. Along with all its other capabilities, this camera features body temperature measurement with an accuracy of ±0.3C in the presence of an adjacent blackbody set at 35 degrees.

The idea of a BTM camera solution is that it allows detection of high body temperature and possible patients can then be retested at a nursing station and encouraged to seek further medical assistance if high temperature is confirmed and there are no mitigating circumstances – things like recent exercise or wearing thermal headgear also tend to raise temperature. Dahua’s BTM solution is being installed in many applications, with 20,000 sold globally, including many in Europe.

It’s tempting to rush ahead to BTM performance but let’s stop a moment and quickly run through the specification. This PoE camera is a dual lens, fixed bullet camera designed to deliver optical video surveillance with Smart IR in applications requiring IP66 ratings, while offering thermal technology to enhance surveillance capabilities, as well as to deliver BTM measurements.

There are a pair of camera systems and a pair of light sources – one is IR, the other white light.

On the thermal side the camera uses uncooled Vox microbolometer technology with a thermal sensitivity of <50mK at an aperture of F1.1 from its fixed lens to ensure the best balance of performance, affordability and compactness. The thermal sensor has a resolution of 256(H)x192(V), while pixel sizes are 12um and spectral range is 8~14um. Thermal colour palettes are selectable, with options including whitehot, blackhot, ironrow, icefire, fusion, rainbow, globow and iconbow1.

The temperature measurement range is from 30C to 45C and the accuracy is ±0.3°C (±0.3 per cent) with black-body in the image and ±1°C (±1 per cent) without a black-body in the scene. Measurement mode can be set to spot, line, or area, and the measurement function supports 12 rules simultaneously.

Alongside its thermal capability, this hybrid camera includes a 1/2.8-inch 2MP progressive scan Sony CMOS, with region of interest, motion detection and colour palettes, active deterrence with white light and siren, 2 alarm inputs and outputs, MicroSD memory and IP67 rating. The optical camera has a varifocal lens with a range from 3.5mm – 7.1mm offering a horizontal angle of view of 50.6 degrees at the wide end and 24 degrees at the long end.

Adjustable image settings include brightness, sharpness, region of interest, auto gain control, FFC and 3D DNR. Video compression options include H.265, H.264 and MJPEG, while the main optical stream is 1080p at 25 images per second and the thermal stream is 1280 x 960 at 25 images per second. There are also thermal and optical sub streams. Bit rate control options include CBR/VBR with H.264 ranging between 640 and 8192 Kbps. Audio compression options include G.711a, G.711Mu and AAC.

Dahua’s 16-input AI NVR is used in this solution. 

The optical camera features day/night functionality and there’s auto (ICR), colour and B/W, BLC modes include BLC/HLC/WDR, white balance options include auto and manual, there’s Ultra1 digital noise reduction, motion detection off/on (4-zone, rectangle) and region of interest off/on (4-zone). Other features include defog, 180-degree flip, mirroring, and privacy masking (4-area, rectangle). A useful feature is active deterrence, which can be used to warn off intruders using integrated and adjustable white light (you can see a white light LED on the camera face) and a siren that can be automated to operate independently of security operators.

This PoE camera features the usual RJ-45 (10/100Base-T) connection and supports the usual range of protocols, including IPv4/IPv6, HTTP, HTTPS, SSL, TCP/IP, UDP, UPnP, ICMP, IGMP, SNMP, RTSP, RTP, SMTP, NTP, DHCP, DNS, PPPOE, DDNS, FTP, IP Filter, QoS, Bonjour and 802.1x. There’s also interoperability with ONVIF Profiles S and G, and API. Streaming is unicast/multicast and the camera supports up to 20 users.

Recording can be via NVR, server-based, NAS, connected workstation/laptop or MicroSD (128GB), while viewing options include Web Viewer IE11, Chrome45, Firefox52, as well as management software options including Smart PSS, DSS, Smart Phone Android, and IOS.

Test Driving the Dahua Thermal Hybrid

I got a demo of Dahua’s body temperature management solution with Dahua’s Simon Cao, who pointed out that the Dahua’s PoE DH-TPC-BF3221-T thermal hybrid bullet camera is designed to help end users in many applications detect high body temperatures that might indicate COVID-19 infection. The procedure is that after elevated temperature is detected, potential patients get a second temperature test before seeking a COVID-19 test should that be required.

The BTM station we’re testing is set up at the Dahua office in Artarmon and it’s quite compact – a camera on a tripod, a thermal blackbody, an NVR, and a laptop to display thermal and optical images. Setup is modular and very simple, and Cao says the system packs into 2 boxes and commissioning takes about half an hour. There’s no need for the system to be connected to the internet and the camera could drive an audible or visual alarm with no monitor.

Simon Cao, Dahua Technologies.

“We must change our behaviours in the face of COVID-19 and Dahua’s PoE DH-TPC-BF3221-T thermal hybrid bullet camera is designed to help end users in many applications do this, with a BTM solution supported by medical procedures,” Cao explains.

As we go along it becomes obvious that managing the system is very easy. Cao says that a key part of the system is the blackbody, which is a heater that’s always at 35 degrees to create a reference point for the camera allowing variations to be established with a high level of accuracy. There’s also an AI NVR, a laptop and a remote monitor in this solution.

“We have an LED light here and a built-in speaker here, so an alarm will activate when the camera detects high temperature – an alarm will beep and the strobe will flash,” Cao says. “What I am going to is hold some hot water beside my face just to give an approximation of the way the system works when it detects temperatures beyond the programmed threshold for an alarm.”

“Importantly, the system will not go into alarm when any elevated temperatures are detected outside the blue background – the system does not recognise anything hot outside that area so getting the background positioned correctly is another part of setup.”

Cao holds the hot water in his hand and moves it around the scene and there’s no reaction from the system. But as soon as he holds the hot water close to his face, the system goes into alarm, with the siren sounding and the strobe flashing. Importantly, these sounds are relatively muted as you’d prefer in an internal environment. It’s also possible to turn the siren off and simply have the strobe flash, depending on client needs.

You can see the reference blackbody sitting at 35C in the top right hand image.

In order to show me this feature a bunch of the Dahua team crowd into the field of view with me and yes, there’s no problem registering 7 subjects.

“The range from camera lens to subject needs to be 3 metres – if the subject gets too close to the lens it will impact on temperature registered,” Cao explains. “And the system can measure the temperatures of multiple subjects simultaneously – 5-7 subjects is easy – the maximum is 15 subjects at a time but given the compact area that’s not easy in normal conditions.

Management of the system is also straightforward.

“You can see on the interface here on the laptop how many people have passed the BTM point and you just double click to see what happened a few moments ago,” Cao says. “The system will replay so operators can check an event. It’s possible to generate daily, monthly and annual reports highlighting the number of people tested, and showing how many had elevated temperatures.

“Our DSS VMS can be used to support the system, or an NVR can be used – I think the NVR is the more flexible option. In this application we are using our Dahua AI NVR because it’s the one that can support the face recognition function and some other features required.

“A good feature with management of the system is the snapshot and the record function – you can see every person’s recorded information along with their temperature. It’s also possible to use a web browser to view the image on a laptop.”

Management of the system is straightforward and historical reporting is simple, too. 

Often the system will be checking the temperatures of staff at a hospital or business as they pass by and security or reception will not need to monitor the screen the entire time – the system will generate an alarm if high temperature is detected. This could be generated by the system, via a laptop connected to the system, or in the VMS if the BTM solution is integrated with it. When high temperature is detected it can then be confirmed at a nursing station and any false positives can be ignored.

“A benefit of BTM solutions is that they are not invasive and subjects don’t need to slow down or stop as they approach the scanning point but can just continue along as they normally would, which keeps pedestrians moving – there’s no need for queuing in busy or crowded environments,” Cao explains.

The thermal blackbody provides a 35C reference point in the field of view.

There has been some conjecture around the accuracy of BTM systems, which may show elevated temperature if a person is wearing warm clothing or has just exercised or come in from outside. Cao says it’s important that a BTM camera solution is seen as a tool that can be used by security teams to direct staff or members of the public to a nursing station, depending on the application.

My impressions of Dahua’s BTM solution and the DH-TPC-BF3221-T camera are of the simple effectiveness of operation and the speed of detection, along with the intuitive nature of Dahua’s interface. It’s all quite straightforward.

Detecting high temperatures using a tumbler of hot water doesn’t make for high science, but as the temperature of the water drops close to human body temperature, it’s possible to observe consistent detection performance from the Dahua solution and to get an idea of the operational requirements of the system. I keep an eye on the thermal blackbody on the monitor – it stays rock steady at 35C throughout the test process.

We’re going to see more many solutions like this over the next couple of years, especially as their prices come down and form factor options expand. Supported by medical procedures, their ability to lower the risk of COVID-19 spread by highlighting a possible patient is beyond question.

Features of the DH-TPC-BF3221-T:

* 256 x 192-pixel VOx uncooled thermal sensor technology
* Athermalized Lens (thermal camera), focus-free
* 1/2.8-inch 2MP progressive scan Sony CMOS
* Support ROI, motion detection, colour palettes
* Support measure body temperature,
* Measurement accuracy: Max ±0.3°C, with blackbody
* Active deterrence with white light and siren
* Built-in 2/2 alarm in/out
* PoE, Micro SD memory, IP67, PoE.