The first smoke detector worked with ionization method. It was long way from these smoke detectors till the video image-based fire detection systems. This way will go towards to better and more improved life- and property safety methods.

The scale of visual methods family is so wide. In this family we can find fire safety as
a young member. Despite, fire safety is young, in 2018 we can find such an extensive arsenal of it. In my diploma thesis I expounded the differences of these video based devices in structural, operational, and in usability for detect different fire attributions. These features don’t allow us to evaluate all of them in the same way. I studied and showed 3 cameras, which I used for fire tests. I summarize the results that I have obtained, in an objective way, stayed in the applied technology.

Cameras, equipped with traditional optics, without any special tools or special softwares, just like my computer’s Sony IMX 175 camera, are the most primitive tools on the list of the tested optics. Most security systems installed for property protection purposes, and they use cameras like Sony IMX 175, or similar, non-fire protection specific cameras.
To program these video tools with appropriate framework environment, and generating
a background database, to be able to use for fire protection, is cumbersome, time consuming, and needs more human and financial resources. It is not worth it; very hard to regain or not even able to regain anything. Moreover, it is not a reliable option.

I used AXIS Q2901-E Temperature Alarm Camera for testing temperature monitoring or heat monitoring cameras. I set up the emissivity to 0,95. I choosed this value with correctional metering with thermocouples and with Fluke 572 Handheld Infrared Thermometer. The fire in the paint box behind the column, located at a distance of 7,5 m, was not detected by any of the cameras. Then I removed the box from the hide by 1-1,5 cm, so the temperature camera was already detecting the temperature difference. After a few seconds, the 12 ℃ temperature appeared on the side of the column. Compared to conventional non-colored non-germanium-based lenses, AXIS was recognized here. It has to be mentioned, when the fire was able to be seen by AXIS, it actually detected the temperature difference within less than 3 seconds.

The results of Bosch AVIOTEC IP Starlight 8000 was proved to be the most forward-looking ones. 13 tests were focused flame tests and 4 were smoke tests out of the total 17.

All of the 7 flame tests performed with Telezoom have been successful. The shortest detection time is 9 seconds only. At the longest detection time, the image of the flame at 10 m is very small for the camera. This was even more complicated because of the cover-up of the camera image. This resulted a cumbersome detection of 43 seconds. However, their average time is not more than 20 seconds. 6 flame tests have been done with Widezoom. Out of these tests, 2 detections were successful with 8 seconds and 15 seconds, with an avarage 12 seconds. Out of the 4 non-observed flames I have covered and uncovered them several times and the flame size was very small. A total of 4 smoke tests have been done; 1 sec, 2 secs, 4 secs and no detection results have been obtained. Their average detection time is 1,75 sec. The only unsuccessful smoke test was because of the background color, with Widezoom with 4,1 mm focal length setting. The fastest smoke detection came from a distance of 5 m.

Summarizing the 17 tests I can say that the alarm signals generated within short time were generated by the Bosch AVIOTEC IP Starlight 8000. Out of the 17 fire tests carried out, no fire attributes were detected on 5 occasions. It means that 70,58 % of the fire detections were effective in the 17 test conditions. The 4 smoke detection tests were 75 % successful and the flame detection tests were 69,23 % successful. During the flame tests, only Widezoom optics with 4,1 mm focal length optics was set to omit detection, but with Telezoom, 9 mm focal length optics, every flame has been detected by the device, which results a 100 % efficiency.

Video-based fire detection cameras can primarily be used to detect fires with high reliability, where suspected fires are associated with prominent flame formation and/or smoke evolution. In the case of flame, for certain devices it is enough to flame in 2 % of the image visible to the camera, but in the case of smoke it is also necessary that the angle of the vertical is not greater than 45° – depends on the device as well. The color of the environment behind the smoke and/or the flame, seen by the video-based fire detector, is also an important part. This greatly influences the success of the detection.

Based on the results, I can conclude that a worthy companion of smoke detectors at the top of the “hit list” of fire detectors (Bellus [2002b]) is a properly and well configured video-based fire detection camera.

I have conducted discussions with 2 staff of the National Disaster Management Directorate, 2 employees of the Budapest Disaster Management Directorate, 1 staff of the Hungarian Chamber of Engineers and 2 employees of MOL Plc.. Based on the knowledge of the listed organizations and organizational units there is no video-based fire detection in Hungary, but leastways not in an authorized and official form. If there is no standard matching, there is no possibility for directing the alert device to send an alert signal to the fire department and to initiate the control. However, based on the results of my tests, they could be used as prefixes or as support for devices that are standardized and authenticated.

The disadvantage of the cameras developed for the detection of fires, is the higher price compared to the more popular “hit products”. However, the extent of the potential damage may be higher on the occasion of a non-observed fire, furthermore, the protection and the preservation of human life cannot be measured in money.


BELLUS, L. [2002B] A tűzjelzés fizikája, a tűzjellemzők érzékelése III. – füst. Tanulmány – elérhető: Védelem online:, kereső:, kulcsszavak: tűzjelző, érzékel, spektrum, letöltés: 2018-04-08

Kapcsolódó cikkek