Image processing techniques are widely used
now a day to detect fire from still images and videos. The
basic technique to label a pixel as fire pixel is by analyzing
its color information. One of the main setbacks of color
models is that they can wrongly detect fire-colored ordinary
objects as fire regions. It is very important to distinguish
between fire and fire-colored ordinary objects, because
generating a lot of false alarms reduces the efficiency of the
fire detection system to a great extent. Various
characteristics of the fire flame other than its color clues are
utilized to make the fire pixel classification more precise.
The fire flame has a distinctive texture characteristic than
other ordinary objects, so combining the texture analysis
along with the color information reduces the false alarms
considerably. Another unique characteristic of the fire flame
is its flickering motion. Proper classification of smoke pixels
also helps to detect the fire at its initial stage. This paper is
a study of different techniques that can drastically reduce
the false positive rate.
Sam G. Benjamin : received his B.Tech (Information Technology)
degree from University of Kerala, Trivandrum in 2010. He is currently
pursuing his Masters in Computer Science & Engineering from
University of Kerala. His research interests focuses on image
processing, data mining, and image mining.
Radhakrishnan B : is working as Asst. Professor in computer
science department. He has more than 14 years experience in
teaching and has published papers on data mining and image
processing. His research interests include image processing, data
mining, image mining.
Rule Based Color Model
Classification of Fire
Pixels
Texture Analysis
Flicker Motion Analysis
Fuzzy
Inference System
All the methods discussed here have good detection rate.
One of the important characteristic that determines the
efficiency of a fire detection system is its false positive
rate. The table 1 compares the false positive rate of the
different methods. Among the various methods discussed
here B. Ugur Toreyin et al. [5] has the least false positive
rate. Table 2 compares the output produced by three
different approaches. It can be clearly seen that using
color information alone wrongly labels fire colored
ordinary objects as fire pixels, thereby generating a lot of
false alarms. Combining color clues, motion analysis and
fire flickering techniques eliminated the false positive rate
to a great extent. Hence for obtaining better results, it is
essential to combine multiple techniques rather than
sticking onto color information alone.
[1] Turgay Celik , Hasan Demirel, “Fire detection in
video sequences using a generic color model”, Fire
Safety Journal (2008).
[2] Turgay Celik, Huseyin Ozkaramanli, and Hasan
Demirel, “Fire and smoke detection without sensors:
Image processing based approach”, 15th European
Signal Processing Conference (EUSIPCO 2007),
Poznan, Poland, September 3-7, 2007.
[3] Daniel Y. T. Chino, Letricia P. S. Avalhais, Jose F.
Rodrigues Jr., Agma J. M. Traina, “BoWFire:
Detection of Fire in Still Images by Integrating Pixel
Color and Texture Analysis”, Proceedings of the 28th
SIBGRAPI Conference on Graphics, Patterns and
Images, 2015.
[4] B. Ugur Toreyin, Yigithan Dedeoglu, A. Enis Cetin,
“Flame Detection In Video Using Hidden Markov
Models”, Bilkent University, TR-06800 Bilkent,
Ankara, Turkey.
[5] B. Ugur Toreyin, Yigithan Dedeoglu, Ugur Gudukbay,
A. Enis Cetin, “Computer vision based method for
real-time fire and flame detection”, Pattern
Recognition Letters 27 (2006) 49–58.
[6] T. Celik, H. Demirel, H. Ozkaramanli, “Automatic
Fire Detection in Video Sequences”, 14th European
Signal Processing Conference (EUSIPCO 2006),
Florence, Italy, September 4-8, 2006, copyright by
EURASIP.
[7] Wen-Bing Horng, Jian-Wen Peng, and Chih-Yuan
Chen, “A New Image-Based Real-Time Flame
Detection Method Using Color Analysis”, IEEE
International Conference on Networking, Sensing and
Control, Tucson, Arizona, USA, March 19-22, 2005.
[8] Punam Patel, Shamik Tiwari, “Flame Detection Using
Image Processing Techniques”, International Journal
of Computer Applications (0975 – 8887) Volume 58–
No.18, November 2012.
[9] Turgay Celik, Huseyin Ozkaramanli, and Hasan
Demirel, “Fire pixel classification using fuzzy logic
and statistical color model”, Acoustics, Speech and
Signal Processing, 2007. ICASSP 2007. IEEE
International Conference on (Volume:1 ).
[10] Peer Neubert and Peter Protzel, “Superpixel
Benchmark and Comparison”, Chemnitz University of
Technology Department of Electrical Engineering and
Information Technology.
[11] Norsyahirah Izzati binti Zaidi, Nor Anis Aneza binti
Lokman, Mohd Razali bin Daud, Hendriyawan
Achmad and Khor Ai Chia, “Fire Recognition Using
Rgb And Ycbcr Color Space”, ARPN Journal of
Engineering and Applied Sciences, VOL. 10, NO. 21,
NOVEMBER 2015.
[12] http://www.peterkovesi.com/projects/segmentation/
[13] https://en.wikipedia.org/wiki/
