DEVELOPING FIRE DETECTION SYSTEM

Jyoti Islam

DEVELOPING FIRE DETECTION SYSTEM

Jyoti Islam

Sign up for access to the world's latest research

checkGet notified about relevant papers

checkSave papers to use in your research

checkJoin the discussion with peers

checkTrack your impact

Abstract

Fire is one of the most serious catastrophic disasters that can strike anywhere and can be very destructive. A method to detect fire would allow the authorities to detect and put out the fires before it becomes out of control. However, most of the available fire detection system uses temperature or smoke sensors which take time to response. Moreover, these systems are costly and not effective if fire is far away from the detectors. This led to think of alternatives such as computer-vision based techniques. One of the cost effective methods would be to use surveillance cameras to detect the fires in order to inform the relevant parties. The proposed project work suggests a method to use surveillance cameras in order to monitor occurrences of fire anywhere within camera range. In this project, two methods are proposed for fire detection in video images using color and motion properties of fire. The first approach uses only color segmentation. The second approach finds the boundary of the moving region in the color segmented image and calculate the amount of fire pixels in this area. Then a fire detection system is developed based on these methods to detect fire efficiently to save life and property from fire hazard.

Figures (67)

FIGURE 1.2: Block Diagram of Proposed Fire Detection System Architecture. Figure 1.2 shows the block diagram of the proposed system architecture.

volatile gases and leave behind the carbon. That is why a charcoal fire burns witl called charcoal. Charcoal is wood that has been heated to remove nearly all of the The actual burning of wood then happens in two separate reactions : (1) When the volatile gases are hot enough(about 260 degrees C for wood), the compound

find whether there is really fire in the moving fire colored region or not. is calculated. Since fire region varies in a great extend, based on these system can As in the proposed system, the color and motion information of fire is used, it is boundary of the moving region is found and the amount of fire pixels in this area

amplitude values of f are all finite, discrete quantities, the image is called a digita A digital image a|m,n] described in a 2D discrete space is derived from an analog image a(x,y) in a 2D continuous space through a sampling process that is fre-

triangle defined by those primary colors [8}.

performed on image of Figure 2.5. FIGURE 2.5: Fire Image in RGB Color Space.

FIGURE 2.6: Image of Fire After RGB to Gray Scale Conversion.

performed on Image of Figure 2.7. FIGURE 2.7: Image before Noise Reduction.

FIGURE 2.8: Image After Noise Reduction.

typically to reduce image noise and reduce detail. FIGURE 2.9: Image Before Smoothing.

Another is for converting gray scale image to bi-level or black-and-white image

FIGURE 2.14: Image Before Contour Detection.

FIGURE 2.16: Image Before Morphological Operation. is one of the options for a full description.

FIGURE 2.18: ] Erosion using Structuring | Hlement S. 2.8.10.2 Dilation

2.9 OpenCV - Open Computer Vision Library

FIGURE 3.1: Block Diagram of Fire Detection System Using Color Segmenta- +10n These steps will be discussed in details below:

chapter 3. Proposed Fire Detection System and Implementation FIGURE 3.2: Fire pixel properties in the RGB color space.

output of system for this scenario. FIGURE 4.1: Indoor Day Environment - No Fire In this scenario there is no fire in the monitored area. Figure 4.1 shows a sample

FIGURE 4.8: Indoor Day Environment - Fire present in indoor day environment.

FIGURE 4.9: Outdoor Day Environment - No Fire fire here. Figure 4.9 shows a sample output of system for this scenario.

FIGURE 4.10: Outdoor Day Environment - Fire 4.10 shows a sample output of system for this scenario.

FIGURE 4.11: Outdoor Day Environment - Paper Fire 4.11 shows a sample output of system for this scenario.

detect fire in this scenario. Figure 4.12 shows correct fire detection in this scenario.

FIGURE 4.14: Night Indoor Environment 1.14 shows a sample output of system for this scenario.

FIGURE 4.15: Moving Camera lJetect it. Figure 4.15 shows a sample output of system for this scenario.

FIGURE 4.16: Night Outdoor Environment 4.16 shows a sample output of system for this scenario.

FIGURE 4.17: Extreme Fire Colored Object system for this scenario.

FIGURE 4.18: Indoor Day Environment - No Fire for this scenario.

Input Image: FIGURE 4.19: Input Image Now the step by step processing of the video frame will be presented.

Background Image: FIGURE 4.20: Background Image

Figure 4.26 shows a sample output of second approach of fire detection when fire is present in indoor day environment.

FIGURE 4.26: Indoor Day Environment - Fire

FIGURE 4.27: Outdoor Day Environment - No Fire ire in this scene. Figure 4.27 shows a sample output of system for this scenario.

FIGURE 4.28: Outdoor Day Environment - Fire 4.28 shows a sample output of system for this scenario.

FIGURE 4.29: Outdoor Day Environment - Paper Fire 4.29 shows a sample output of system for this scenario.

detect fire in this scenario. Figure 4.30 shows correct fire detection in this scenario.

fire detection sometimes fails in this scenario.

FIGURE 4.32: Night Indoor Environment |.32 shows a sample output of system for this scenario.

FIGURE 4.33: Moving Camera letect it. Figure 4.33 shows a sample output of system for this scenario.

FIGURE 4.34: Night Outdoor Environment .34 shows a sample output of system for this scenario.

FIGURE 4.35: Extreme Fire Colored Object formance in this scenario. Figure 4.35 shows a sample output of system for this

TABLE 4.1: System Performance Using Approach One

less false alarm in case of extreme fire colored object.

Table 4.3 shows average frame processing time for the two approaches. Figure 4.3¢

FIGURE 4.36: Average Processing Time Variance for Two Approaches fire detection system.

Table 4.4 shows average detection rate for the two approaches. Figure 4.37 shows

FIGURE 4.37: Average Detection Rate for Two Approaches fire detection system.

Key takeaways

It should be able to detect fire in large open spaces, so that the whole scenario must be considered, and not just a single point on the image from the video feed.
Color and motion information of the image is analyzed in almost every fire detection system.
The image shown in Figure 2.3 represents digitization of a fire image in night environment.
Then according to the detection method, fire presence in the image or not is decided.
Image and video based systems can be an alternative to particle sensors for fire detection.

Gaurav Patil

International Journal of Engineering Research and, 2017

The process of oxidation of any material in the exothermic process of combustion, releasing heat and light as byproducts, is called Fire. The light parameter and the color of the flame helps in detecting fire. Fire detection using color information has many applications in computer vision and other domains. Our color model based method used for fire detection has many advantages over conventional methods of smoke detection etc., such as simplicity, feasibility and understandability. In order to enhance the performance parameters of fire flame detection based on a live video stream, we propose an effective color model based method for fire detection. Eeach and every pixel is checked for the presence or absence of fire using color features, and periodic behavior in fire regions is also analyzed. Dynamic boundary check is also done to detect the edges of the fire Region of Interest (ROI). Candidate fire regions are detected using the chromatic and dynamic measurements.

Log In

DEVELOPING FIRE DETECTION SYSTEM

Sign up for access to the world's latest research

Abstract

Key takeaways

Related papers

Related topics