Real-time Fire and Flame Detection in Video*

B. Uğur Töreyin, Yiğithan Dedeoğlu, Uğur Güdükbay, A. Enis Çetin

Bilkent University, Ankara, Turkey

Conventional point smoke and fire detectors are widely used in buildings. They typically detect the presence of certain particles generated by smoke and fire by ionization or photometry. Alarm is not issued unless particles reach the sensors to activate them. Therefore, they cannot be used in open spaces and large covered areas. Video based fire detection systems can be useful to detect fire in large auditoriums, tunnels, atriums, etc. The strength of using video in fire detection makes it possible to serve large and open spaces. In addition, closed circuit television (CCTV) surveillance systems are currently installed in various public places monitoring indoors and outdoors. Such systems may gain an early fire detection capability with the use of a fire detection software processing the outputs of CCTV cameras in real time.


Our aim is to detect flame, fire and smoke in real-time by processing the video data generated by an ordinary camera monitoring a scene.

Detection Algorithm

The proposed video-based fire detection algorithm consists of four steps:


1. Moving region detection:

Moving pixels or regions in the current frame of a video are determined by using the hybrid background estimation method developed by Collins et al. (1999).

2. Detection of fire-colored pixels

Color values of moving pixels are compared with a pre-determined color distribution, which represents possible fire colors in video in RGB color space.

3. Temporal wavelet analysis

It is observed that turbulent flames flicker with a characteristic frequency of around 10 Hz independent of the burning material and the burner (Albers and Agrawal, 1999; Chamberlin and Rose, 1965). To capture 10 Hz

flicker, the video should capture at least 20 fps. If the video is available at a lower capture rate, aliasing occurs but flicker due to flames can still be observed in the video.


Each pixel’s red channel value, xn[k,l], is fed to a two stage-filter bank (Fig.1). This filter bank is composed of half-band high-pass and low-pass filters with filter coefficients {-0.25, 0.5, -0.25} and {0.25, 0.5, 0.25}, respectively. The filter bank produces wavelet subsignals dn[k,l] and en[k,l]. If there is high frequency activity at pixel location [k,l], high-band subsignals d and e get non-zero values (Fig.2). However, in a stationary pixel, the values of these two subsignals should be equal to zero or very close to zero because of high-pass filters used in subband analysis (Fig.3).


Fig.1. A two-stage filter bank. HPF and LPF represent half-band high-pass and low-pass filters, with filter coefficients {-0.25, 0.5, -0.25} and {0.25, 0.5, 0.25}, respectively. This filter bank is used for wavelet analysis.



Fig.2. (a) Temporal variation of image pixels xn[111, 34] in time. The pixel at [111, 34] is part of a flame for image frames xn, n=1, 2, 3, 19, 23, 24, 41 and 50. It becomes part of the background for n = 12,..., 17, 20, 21, 26, 27, 31,..., 39, 45, 52,..., and 60. Wavelet domain subsignals (b) dn and (c) en reveal the fluctuations of the pixel at [111, 34].



Fig.3. (a) Temporal history of the pixel [18, 34] in time. It is part of a fire-colored object for n = 4, 5, 6, 7, and 8, and it becomes part of the background afterwards. Corresponding subsignals (b) dn and (c) en exhibit stationary behavior for n>8.




4. Spatial wavelet analysis


In an ordinary fire-colored object there will be little spatial variations in the moving region as shown in Fig.4(a). On the other hand, there will be significant spatial variations in a fire region as shown in Fig.5(a). The spatial wavelet analysis of a rectangular frame containing the pixels of fire-colored moving regions is performed. The images in Figs.4(b) and 5(b) are obtained after a single stage two-dimensional wavelet transform that is implemented in a separable manner using the same filters explained in the previous section. Absolute values of low-high, high-low and high-high wavelet subimages are added to obtain these images. A decision parameter is defined for this step, according to the energy of the wavelet subimages.




Fig.4. (a) A child with a fire-colored t-shirt, and (b) the absolute sum of spatial wavelet transform coefficients, |xlh[k,l]| + |xhl[k,l]|+|xhh[k,l]|, of the region bounded by the indicated rectangle.


Fig.5. (a) Fire, and (b) the absolute sum of spatial wavelet transform coefficients, |xlh[k,l]| + |xhl[k,l]| + |xhh[k,l]|, of the region bounded by the indicated rectangle.




* Full version of this text can be found in the article “Computer vision based method for real-time fire and flame detection”, Pattern Recognition Letters 27 (2006) 49-58.