Fire Detection based on Color, Shape and Motion
0 likes56 views
This document describes a computer vision-based fire detection system that analyzes video frames captured by a camera. It uses algorithms to detect fires based on color, shape, and motion in the video frames. The system sends SMS alerts to notify emergency services if a fire is detected. It works by continuously capturing frames with a camera and applying fire detection algorithms to analyze the color, convert the images to HSV color space, and perform background subtraction to isolate moving objects. If fire is detected in the image processing, a microcontroller sends an SMS using a GSM module and also triggers an alarm with a buzzer and message on an LCD display. The system provides faster response times than traditional smoke detectors through real-time video analysis.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 07 | July -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1806
surveillance cameras. . Works on continuous frames of
images are captured by camera. Thus it has faster response
time. These images are monitored by software. After
detecting fire we can manually send message to near by fire
emergency services by GSM module. The advantage is that
the overall performance of thesystemsignificantlyincreases
as compared to traditional smoke detector.
REFERENCES
[1] MA Shu-guang Construction of Wireless Fire Alarm
System Based on ZigBeeTechnology The5thConference
on Performance-based Fire and Fire Protection
Engineering2011 .
[2] Jareerat Seebamrungsat , “ Fire detection in the
buildings using image processing” Student Project
Conference (ICT-ISPC),2014 ThirdICTInternational 26-
27 March 2014 .
[3] Kausik sen ET AL, “Automated Fire Detection and
Controlling System” International Advanced Research
Journal in Science, Engineering and Technology, May
2015.
[4] Pasquale Foggia, Alessia Saggese, “,Real-time Fire
Detection for Video Surveillance Applications using a
Combination of Experts based on Color, Shape and
Motion” 2015.2392531, IEEE .
1](https://image.slidesharecdn.com/irjet-v4i7383-170907100705/85/Fire-Detection-based-on-Color-Shape-and-Motion-3-320.jpg)
Fire Detection based on Color, Shape and Motion
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 07 | July -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1804 Fire Detection Based on Color, Shape and Motion Monika Waghmare1, Prof. Meenakshi Annamalai2 1Department Of Electronics & Telecommunication , JSPM’s Bhivarabai Sawant Institute of Technology and Research, Wagholi Pune, Maharashtra, India 2 Professor, Department Of Electronics & Telecommunication, JSPM’s Bhivarabai Sawant Institute of Technology and Research, Wagholi ,Pune, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - It is method which is able to detect fires by analyzing the videos acquired by cameras . It is the computer vision based fire detection algorithm. Works on continuous frames of images are captured by camera. Thus it has faster response time. These images are monitored by software. After that, fire detection algorithms are applied onthevideosuchas color blurring, RGB to HSV conversion using matlab . Key Words: Camera, MicrocontrollerLPC2148, PC,GSM module, Buzzer,LCD-Display,USBtoserial converter. 1.INTRODUCTION Nowadays different types of fire and smoke detectors are existing in the market. Smoke detectors are used to detect smoke which tells that fire is existing. Several methods have been proposed, used to analyze the videos acquired by traditional video surveillance cameras and detect fires or smoke. There are two main types depending on the analyzed features: color based and motion based and based on the consideration that a flame. In many proposed method, only physical devices like temperature sensor, smoke sensor are used to detect fire. But sometimes it gives false alarms. Hence the computer vision based fire detection algorithm is needed. This technique is used along with conventional fire detection method. In this we study, continuous frames of images arecapturedbycamera.Thus it has faster response time. These images are monitored by software. After that, fire detection algorithms are appliedon the video such as color blurring, RGB to HSV conversion using matlab . The detection of fire by video is mostly appropriate in industrial monitoring and surveillance used to display buildings and environment as a part of an initial warning mechanism that gives message of start of fire. Video-based systems can sense fires at an early stage before tragedy. Conventional fire detectors are costly and they are not suitable for open spaces such as forest . They highly produce false alarms and often detect fire when it’s too late. So we have to design low cost fire detection system with small response time. Computer vision based techniqueshas a great capacity to meet these requirements. This detection of fire in video is mostly used in industrial monitoring , buildings and in forest . Cameras can be attached everywhere freely, therefore detection range is larger. Technology has higher consistency and real-time performance. 2 METHODOLOGY 2.1 System Overview The system gives the schematic design of fire detection system. Here we have two architectures. One is image processing technique andtheotherishardwarebased conventional method. Here in this module the video of the fire is captured by collecting the sequential frames using camera. We have analyzed theimages whichconsistoffire or smoke samples. These frames of images are processed for detection of fire. This is done usingvariousimageprocessing algorithms. The software based result send to the microcontroller. Serial to USB converter is used as the interface between the PC and microcontroller. Here the SMS will be sending to the firedepartmentthrougha GSMmodem connected to the system using Attention commands. Any abnormality is detectedthenthemicrocontroller will turn on the buzzer. When detects fire switch on a water sprayer to spray water. Fig1. Block Diagram of system
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 07 | July -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1805 2.2 Image Processing Image Acquisition: Image acquisition is first step inimage preprocessing. Camera gives RGB data having 640 x 480 resolutions. Camera gives RGB images at the rate of 30 frames per second. Image Pre-processing: After acquisition of image some pre-processing is done on acquired image. Preprocessing includes noise filtering using median filter. Noise Filtering: Noises in the imagewill degradethe accuracy of image. In our project we are removing ‘Paper and Salt Noise’. This noise can be removed by using Median Filter. Median Filtering: Median filtering is used to reduce “salt and pepper” noise. The Median Filter replaces the central estimation of M-by-N neighborhood with its middle value. Background is maintained and properly updated (Background updating) so it changes of the environments during the day; then, a background subtraction strategy is applied in order to obtain the foregroundmask,encodingthe objects moving in the scene (Foreground mask extraction). We get , the blobs, each one being associated to an object . Fig.2. Image processing Part 3. IMPLEMENTATION 3.1 Hardware and Software used 1 Camera 2Power supply 5v 3ARM7 4 LCD display 5 USB to serial converter 6 GSM module 7 Matlab 2013a(software) 4.RESULT 4.1 Output using GUI Using MATLAB coding we created GUI. We detect fire usingcameraandbyconnectingCOMportto PC. We entered registered mobile number on which message would be sent in case of fire detection. Fig -3: Snapshot of GUI 4.2 Output on LCD display and buzzer Whenever fire is detected we will get message on LCD. Display that fire is detected and buzzer will start ringing. Fig -4: LCD display on circuit 5. CONCLUSIONS This project discussed about a method whichis able to detect fires by analyzing the videos acquired by
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056 Volume: 04 Issue: 07 | July -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1806 surveillance cameras. . Works on continuous frames of images are captured by camera. Thus it has faster response time. These images are monitored by software. After detecting fire we can manually send message to near by fire emergency services by GSM module. The advantage is that the overall performance of thesystemsignificantlyincreases as compared to traditional smoke detector. REFERENCES [1] MA Shu-guang Construction of Wireless Fire Alarm System Based on ZigBeeTechnology The5thConference on Performance-based Fire and Fire Protection Engineering2011 . [2] Jareerat Seebamrungsat , “ Fire detection in the buildings using image processing” Student Project Conference (ICT-ISPC),2014 ThirdICTInternational 26- 27 March 2014 . [3] Kausik sen ET AL, “Automated Fire Detection and Controlling System” International Advanced Research Journal in Science, Engineering and Technology, May 2015. [4] Pasquale Foggia, Alessia Saggese, “,Real-time Fire Detection for Video Surveillance Applications using a Combination of Experts based on Color, Shape and Motion” 2015.2392531, IEEE . 1