The need for more efficient and effective traffic management systems is increasing as urbanization and population growth continue. The use of thermal imaging as part of a comprehensive traffic management system could provide operators with improved accuracy and efficiency.
Izvor: a&s International
E-mail: redakcija@asadria.com
When used in conjunction with other sensors in a comprehensive traffic management system, thermal imaging sensors, such as thermal cameras, have a lot to offer in terms of accuracy and efficiency.
Although thermal imaging has a long history in military applications and in perimeter detection for physical security, and its use in traffic management is not yet ubiquitous, the use of thermal sensors across verticals is growing rapidly.
A report from MarketsandMarkets estimates the global thermal imaging market will grow from US$7.7 billion in 2016 to $10.3 billion by 2021, at a CAGR of 5.9 percent. Growth is attributed to the rising demand of thermal cameras as a result of declining prices. Furthermore, growth is also being credited to the increasing need of thermal cameras in all commercial applications, including transportation.
In fact, the use of thermal imaging in the automotive application segment is expected to gain traction in the next five years, according to the report. “The increasing use of thermal cameras in the automotive sector to reduce the risks in driving, especially during the night, is one of the major reasons for the growth of thermal imaging market in this application,” stated MarketsandMarkets.
While applying thermal imaging directly to automotive use isn’t the same as using it for traffic management, the need for it in the transportation sector can be inferred. Additionally, with the need for traffic management growing, combining different technologies, including thermal imaging, will only help create more effective solutions.
Why Use Thermal
One of the biggest draws of thermal imaging cameras across sectors is their ability to “see” in the dark and inclement weather. They can also measure the temperature of any object in their field of view, allowing them to detect fires at an early stage over the full detection range.
“This unlike other fire detection technologies requires no contact with flames or heated gasses, nor is any smoke propagation needed for the camera to detect excessive heat generated by fire or another vehicle malfunction,” explained Michael Deruytter, Director of Innovation at FLIR Intelligent Transportation Systems. “An additional benefit of thermal imaging is that this technology enables operators to see through smoke. This can be a life-saving feature in smoke-filled tunnels and can provide valuable information to firefighting teams about the possible location of people.”
Deruytter further explained that because thermal imaging cameras do not make use of visible light, but rely on thermal radiation or heat given off by everything in their field of view, they do not get confused by sun glare, darkness, headlights, shadows, wet streets, snow or fog. This makes them an ideal component of a 24/7 traffic monitoring solution.
Thermal and Traffic Analytics
Thermal imaging on its own is not a complete solution for traffic management. However, when applied with video analytics, thermal imaging cameras can be very beneficial.
“Thermal cameras produce images similar to the standards rendered by visible cameras, especially in terms of the shape of the object. This means that video analytics developed for visible cameras to conduct shape recognition, motion detection, intrusion detection, people counting, etc., can also be applied to thermal image cameras,” said Emmanuel Bercier, Strategic Marketing Manager at ULIS, a France-based manufacturer of thermal imaging sensors. “Moreover, since thermal cameras detect heat from objects, objects are highly distinguishable from their surroundings. As a consequence, this will simplify existing video analytics as it eliminates the need to take light variations or shadow effects into account,” Bercier continued.
Deruytter pointed out that thermal imaging cameras are well suited for video analytics use since the pixels in the image represent heat information — the hotter an object, the more thermal energy is being radiated; however, even extremely cold objects, up to negative 243°C (zero Kelvin), radiate thermal energy.
“Thermal cameras can detect the heat given off by cars, cyclists and pedestrians. This detection information can be used to classify the object or road user and make smart decisions for intersection control in seconds,” Deruytter said. “Intelligent thermal cameras are also used to detect incidents, such as stopped vehicles or wrong-way drivers, which, in turn, speeds up intervention by traffic authorities and emergency services.”
Challenges for Thermal in Traffic
Though thermal cameras do provide better detection, they will mainly be used in conjunction with visible cameras. Combining both camera types will ultimately provide the best detection performance and support identification, for example identifying a license plate number,” Bercier said.
Since both thermal and visible-light technologies have their own unique strengths, combining them makes perfect sense.
“This multi-sensor approach has been used in high-end surveillance and border security applications, where a combination of visible and thermal cameras identify and address a wide variety of security threats,” Deruytter explained. “In traffic applications, combining these two camera technologies into one system enables traffic authorities to use the superior detection abilities of thermal cameras while gaining important details with HD visible-light cameras.”
Cost is another factor holding end users and systems integrators back from deploying thermal cameras for traffic management. “Over the last decade, costs have been greatly reduced.
One can now find affordable thermal cameras for traffic management in the range of a few hundred dollars,” Bercier said. He added that thermal cameras now resemble commodity cameras in that they come with standard protocol and interfaces.
“This is positive in making it easier to integrate them in new or existing infrastructures. We are increasingly seeing video management systems integrators consider thermal camera equipment. So infrastructures are almost ready to embed thermal imaging technology,” Bercier added.
Traffic to Bring on the Heat
With the price of thermal continuing to decline and the need for accurate traffic management on the rise, more and more systems will incorporate thermal imaging technology into its solutions. As it was pointed out, thermal imaging by itself is not ideal for traffic management, but as a part of a bigger solutions that utilizes visible-light cameras and video analytics, thermal imaging will bring an extra layer of accuracy and efficiency to traffic management.
While visible-light cameras are most commonly used for traffic management, thermal cameras could provide benefits that others can’t.
Visible-light cameras have historically been very popular for traffic monitoring. This, according to Michael Deruytter, Director of Innovation at FLIR Intelligent Transportation Systems, is because when a traffic event occurs, it provides immediate feedback to an operator in the traffic management center to take futher actions.
However, since visible cameras are dependent on the light reflected by objects to produce an image, the reliability of their performance and capacity to detect is extremely sensitive to outside light conditions.
“The majority of cameras used in video surveillance or traffic management are sensitive to light in the visible spectrum, typically 0.4 to 0.8 µm wavelength or up to 1.0 for cameras sensitive in the near infrared. Visible cameras are limited to detecting only those objects exposed to sunlight or an external light source, such as street lighting,” explained Emmanuel Bercier, Strategic Marketing Manager at ULIS.
“Whether an object is human, an animal, a vehicle, a road or tree, it will emit heat energy depending on two elements: its inherent temperature and its material substance. All objects above zero kelvins radiate heat emitted in the far infrared spectral bandwidth 8 to 12µm. Thermal imaging cameras are based on a technology that is sensitive in far infrared spectrum, which means they are only sensitive to the light directly emitted by an object,” he added.
Thus, thermal imaging cameras are able to provide constant visibility and reliable object detection, 24/7, whatever the light, atmospheric or environmental conditions, such as fog or smoke. As such, thermal cameras are well suited for traffic management which is subject to the ever-changing nature of outside elements.
Despite the many benefits of using thermal imaging for traffic management, not all companies believe it is the right technology for the job.
Thermal imaging technology is being used more and more for traffic management, but its use doesn’t necessarily mean it is the best fit. In fact, Daniel Chau, Overseas Marketing Director at Dahua Technology, explained why thermal imaging cameras are not Dahua’s recommended products for traffic management. The company, which has its own line of thermal imaging cameras, mainly applies its products in perimeter protection, forest fire detection and monitoring of high-voltage power transformer substations.
Chau notes that the key advantage of thermal cameras in traffic management is the higher accuracy rate in counting passing vehicles; however, the 5- to 10-percent improvement rate is marginal when compared to good visible-light cameras with artificial intelligence (AI) analytics. Furthermore, it does not bring “material value to the outcome of traffic management.”
“AI and deep learning are enabling visible-light imaging surveillance solutions to recognize vehicle details such as car color, plate number, make, model and use of seat belt. This information could bring significant value to traffic management by understanding the composition of traffic at different times of day,” Chau explained.
“The advantage of thermal cameras does not justify its higher price nor its lack of intelligent recognition and analytics capabilities, thus thermal cameras are not widely applied in traffic management,” he added. Instead, Chau believes AI and deep-learning-powered visible-light imaging surveillance that can deliver advanced analytics function such as video synopsis, human recognition and vehicle recognition are more suitable for traffic management than thermal imaging surveillance.
