Fast and Efficient Method for Fire Detection Using Image Processing

Turgay Celik 한국전자통신연구원 2010 ETRI Journal Vol.32 No.6

Conventional fire detection systems use physical sensors to detect fire. Chemical properties of particles in the air are acquired by sensors and are used by conventional fire detection systems to raise an alarm. However, this can also cause false alarms; for example, a person smoking in a room may trigger a typical fire alarm system. In order to manage false alarms of conventional fire detection systems, a computer vision-based fire detection algorithm is proposed in this paper. The proposed fire detection algorithm consists of two main parts: fire color modeling and motion detection. The algorithm can be used in parallel with conventional fire detection systems to reduce false alarms. It can also be deployed as a stand-alone system to detect fire by using video frames acquired through a video acquisition device. A novel fire color model is developed in CIE L*a*b* color space to identify fire pixels. The proposed fire color model is tested with ten diverse video sequences including different types of fire. The experimental results are quite encouraging in terms of correctly classifying fire pixels according to color information only. The overall fire detection system’s performance is tested over a benchmark fire video database, and its performance is compared with the state-of-the-art fire detection method.

Fire Detection System using Faster R-CNN

Young-Jin Kim,Eun-Gyung Kim 한국정보통신학회 2017 2016 INTERNATIONAL CONFERENCE Vol.9 No.1

Various fire detection systems have been constructed to prevent disastrous fire. However, existing fire detection systems are limited to practical applications due to lower detection accuracy and frequent alerts caused by incorrect operations. Previous fire detection systems have only focused on detecting flames. Therefore they can mistake the flames of candles or gas ranges as a fire. They also cannot provide additional life-saving information, such as the location of people or fire extinguishers. Thus, we have tried to construct a new fire detection system which can improve flame detection accuracy, does not incorrectly identify the flame of candles or gas ranges as a fire, and also provide additional lifesaving information. Faster R-CNN is a deep learning algorithm that detects classes and locations of objects, as well as fires, in real-time by using CNN. We have built our fire detection system based on Faster R-CNN. In order to evaluate the performance of our fire detection system, we used various images such as forest fires, gas range fires, and candle flames. Consequently, the fire detection rate of our system was very good at 99.24%. In addition, we analyzed its object detection performance involving 14 classes, such as people, fire extinguishers, doors, pets, etc. Finally, the mAP (mean Average Precision) was relatively high at 0.7863.

Study on the improvement of Malfunction of automatic fire detection system on the urban rail

Jun-su Kim,Young-hwan Jang,Cheon-heon An,Chung-su Kim 한국도시철도학회 2016 IJAR Vol.4 No.1

If automatic fire detection equipment which is consist of fire protection system which detected in the early stages of a fire and interlocked other fire protection equipment alarm a false alarm (there are no fire, but it alarm there are fire), cause confuse. If these kinds of error occur again and again, even manager can be turn fire detection equipment off. Because there frequently occur false alarms, so turn fire detection equipment off, if really a fire occur, the fire detector can’t alarm and interlock other fire detection equipment. So it can delay fire suppression, evacuation, and produce loss of lives and properties. That’s the reason automatic fire detection equipment is most important in Metro fire system. As public transportation system, metro and its fire detection system notice false alarms cause massive disruption to the public. And it makes loss trustfulness about metro fire detection system, so that the public can’t behave a right response. Therefore, this paper, Metro Fire Department fire sensors could malfunction occurs in the system as reviewed and discussed ways to improve the environment in which public transportation automatic fire detection equipment reliability and attempts to improve. If the automatic fire detection system which detects the fire in the early stage of fire and makes each interlocked fire-fighting equipment to work activates a false fire alarm, it could cause confusion by operation and alarming of fire-fighting equipment. If these false alarms activate repeatedly, a manager would shut the power of receiver down. If the power of receiver is shut off due to frequent false alarming, in case of actual fire, the alarming would not transmit correctly and the interlocked equipment could not work properly. This delay of instant evacuation and initial extinguishment could cause losses of life and property. Therefore the function of automatic fire detection equipment is so important in the fire-fighting system of urban rail. If an abnormal alarming occurs in a railroad as a public traffic method, it could cause big confusion to citizens using the railroad, and citizen’s ability to react would drop noticeably because they don’t give trust to the fire protection system. Therefore this thesis would increase the reliability of automatic fire detection system of public transportation by reviewing the environment which the false alarming could occur in the urban fire protection system, and studying the improvement proposal.

영상기반 자동 화재감지시스템의 감지성능 특성 분석 연구

구재현,Ku. Jae-Hyun 한국방재학회 2013 한국방재학회논문집 Vol.13 No.5

기존의 연기감지기와 불꽃감지기는 화재 발생시 주변 환경의 영향에 의해 감지성능이 크게 제한되어 명확한 화재감지에 문제점을 가지고 있다. 본 연구에서는 화재감지 성능을 향상시키기 위하여 CCTV 카메라, 영상화재수신기, CPU로 구성된 영상기반 자동 화재감지시스템의 감지성능을 분석하고자 한다. 지능형 모션 감지를 위한 GMM(Gaussian Mixture Model) 모델링이 해석되었고 MHI(Motion History Image) 기술의 적용은 화재감지 성능을 향상시킴을 보여 주었다. 화재감지 성능은 화원과 감지거리에 따라 실험적으로 평가되었다. 결과적으로, 화재감지 시간은 감지거리가 증가할수록 증가하며 감지거리 15 m에서 10초로 분석되었다. 따라서, 개발된 영상기반 자동 화재감지 시스템은 불꽃감지기의 감지시간기준 30초와 비교하여 자동 화재감지를 위한 양호한 성능을 보여 주었다. Smoke detectors and flame detectors in detection ability of sensors is greatly limited by the ability to detect environmental impacts in case of fire have a problem in accurate fire detection. This study describes to analyze detection performance of a video-based automated fire detection system consisted of CCTV camera, a video-based fire alarm control station, CPU in order to enhance the fire detection performance. GMM(Gaussian Mixture Model) modeling for an intelligent motion detection is analyzed and the application of MHI(Motion History Image) technology is showed the improved fire detection performance. The fire detection performance is experimentally evaluated as a function of the fire source and detection distance. As a result, the fire detection time is found to increase with increased the detection distance and is analyzed with 10 sec at the detection distance of 15 m. Therefore, the video-based automated fire detection system developed showed good performance for automatic fire detection comparing with the detection time criteria of 30 sec for flame detectors.

사례 분석을 통한 IoT 기반 화재탐지시스템의 화재 감지신호 특성

박승환,김두현,김성철,Park, Seung Hwan,Kim, Doo Hyun,Kim, Sung Chul 한국안전학회 2022 한국안전학회지 Vol.37 No.3

This study aims to provide a fundamental material for identifying fire and no-fire signals using the detection signal characteristics of IoT-based fire-detection systems. Unlike analog automatic fire-detection equipment, IoT-based fire-detection systems employ wireless digital communication and are connected to a server. If a detection signal exceeds a threshold value, the measured values are saved to a server within seconds. This study was conducted with the detection data saved from seven fire accidents that took place in traditional markets from 2020 to 2021, in addition to 233 fire alarm data that have been saved in the K institute from 2016 to 2020. The saved values demonstrated variable and continuous VC-Signals. Additionally, we discovered that the detection signals of two fire accidents in the K institution had a VC-Signal. In the 233 fire alarms that took place over the span of 5 years, 31% of smoke alarms and 30% of temperature alarms demonstrated a VC-Signal. Therefore, if we selectively recognize VC-Signals as fire signals, we can reduce about 70% of false alarms.

An Intelligent Fire Detection Algorithm for Fire Detector

Hong, Sung-Ho,Choi, Moon-Su The Korean Society of Safety 2012 International Journal of Safety Vol.11 No.1

This paper presents a study on the analysis for reducing the number of false alarms in fire detection system. In order to intelligent algorithm fuzzy logic is adopted in developing fire detection system to reduce false alarm. The intelligent fire detection algorithm compared and analyzed the fire and non-fire signatures measured in circuits simulating flame fire and smoldering fire. The algorithm has input variables obtained by fire experiment with K-type thermocouple and optical smoke sensor. Also triangular membership function is used for inference rules. And the antecedent part of inference rules consists of temperature and smoke density, and the consequent part consists of fire probability. A fire-experiment is conducted with paper, plastic, and n-heptane to simulate actual fire situation. The results show that the intelligent fire detection algorithm suggested in this study can more effectively discriminate signatures between fire and similar fire.

화재 탐지 영역의 이미지와 동영상 인식 사이 인공지능 모델 성능 비교 연구

이정록,정상,정서현,이대웅 한국재난정보학회 2023 한국재난정보학회 논문집 Vol.19 No.4

Purpose: We would like to confirm that the false positive rate of flames/smoke is high when detecting fires. Propose a method and dataset to recognize and classify fire situations to reduce the false detection rate. Method: Using the video as learning data, the characteristics of the fire situation were extracted and applied to the classification model. For evaluation, the model performance of Yolov8 and Slowfast were compared and analyzed using the fire dataset conducted by the National Information Society Agency (NIA). Result: YOLO's detection performance varies sensitively depending on the influence of the background, and it was unable to properly detect fires even when the fire scale was too large or too small. Since SlowFast learns the time axis of the video, we confirmed that detects fire excellently even in situations where the shape of an atypical object cannot be clearly inferred because the surrounding area is blurry or bright. Conclusion: It was confirmed that the fire detection rate was more appropriate when using a video-based artificial intelligence detection model rather than using image data.

An Efficient Fire Detection Method Based on Orientation Feature

Mao Ye,Tao Li,Feng Pang,Haiyang Wang,Jian Ding 제어·로봇·시스템학회 2013 International Journal of Control, Automation, and Vol.11 No.5

This paper proposes a novel method for reliable fire detection. The burning fire usually causes rich moving features in terms of directions, which can offer the best chance to distinguish between the fire region and the non-fire one. Motivated by this observation, we design a novel orientation feature to represent this characteristic. Based on this feature, a method is proposed to detect the fire efficiently. First, fire color is utilized to extract the fire candidate areas from the surveillance video. Then, the direction is obtained by computing the optical flow for each pixel in the candidate area. The directions are discretized to four parts. By counting the percentage of pixels whose moving directions fall into these four parts in a period of time, and combining with the two parameters, i.e., both of the number of frames without the moving directions and the number of consecutive frames in the candidate area, we use these six parameters as the fire orientation feature. In the end, by training a support vector machine (SVM) classifier with the input of our fire orientation feature, the candidate area is judged whether it is a fire. Our main contribution is that we design the novel fire orientation feature. The fea-ture can not only characterize the fire intrinsic dynamic properties accurately but also is very efficient. Compared with the art-of-state methods, the experimental results confirm that our approach signifi-cantly improves the accuracy of fire detection and impressively decreases the false alarm rate. The de-tection speed of our approach is also very competitive with the art-of-state fire detection methods.

Fire Detection Approach using Robust Moving-Region Detection and Effective Texture Features of Fire

트룩 뉘엔(Truc Kim Thi Nguyen),강명수(Myeongsu Kang),김철홍(Cheol-Hong Kim),김종면(Jong-Myon Kim) 한국컴퓨터정보학회 2013 韓國컴퓨터情報學會論文誌 Vol.18 No.6

본 논문은 그레이레벨히스토그램을 이용한 움직임 영역검출, 퍼지 클러스터링을 이용한 칼라 분할, 그레이 레벨 동시발생 행렬을 이용한 특징 추출 및 서포터 벡터 머신을 이용한 화재 분류 등과 같은 다중 이종 알고리즘을 포함하고 있는 효과적인 화재 감지 방법을 제안한다. 제안한 방법은 움직임 영역을 검출하기 위해그레이레벨히스토그램에 기초한 최적의 임계값을 결정하고 난 후, CIE LAB 칼라 공간에서 퍼지 클러스터링을 적용하여 칼라 분할을 수행한다. 이러한 두 단계는 화재의 후보 영역을 기술하는데 도움이 된다. 다음으로 그레이 레벨 동시발생 행렬을 이용하여 화재의 특징을 추출하고, 이러한 특징들은 화재인지 아닌지를 분류하기 위해 서포터 벡터 머신의 입력으로 사용된다. 제안한 방법을 평가하기위해 기존의 두 알고리즘과 화재 검출율 및 오류 화재 검출율에서 비교하였다. 모의실험결과, 제안한 방법은 97.94%의 화재 검출율 및 4.63%의 오류 화재 검출율을 보임으로써 기존의 화재 감지 알고리즘보다 우수성을 보였다. This paper proposes an effective fire detection approach that includes the following multiple heterogeneous algorithms: moving region detection using grey level histograms, color segmentation using fuzzy c-means clustering (FCM), feature extraction using a grey level co-occurrence matrix (GLCM), and fire classification using support vector machine (SVM). The proposed approach determines the optimal threshold values based on grey level histograms in order to detect moving regions, and then performs color segmentation in the CIE LAB color space by applying the FCM. These steps help to specify candidate regions of fire. We then extract features of fire using the GLCM and these features are used as inputs of SVM to classify fire or non-fire. We evaluate the proposed approach by comparing it with two state-of-the-art fire detection algorithms in terms of the fire detection rate (or percentages of true positive, PTP) and the false fire detection rate (or percentages of true negative, PTN). Experimental results indicated that the proposed approach outperformed conventional fire detection algorithms by yielding 97.94% for PTP and 4.63% for PTN, respectively.

An Intelligent Fire Detection Algorithm for Fire Detector

( Sung Ho Hong ),( Moon Su Choi ) 한국안전학회(구 한국산업안전학회) 2012 International Journal of Safety Vol.11 No.1

This paper presents a study on the analysis for reducing the number of false alarms in fire detection system. In order to intelligent algorithm fuzzy logic is adopted in developing fire detection system to reduce false alarm. The intelligent fire detection algorithm compared and analyzed the fire and non-fire signatures measured in circuits simulating flame fire and smoldering fire. The algorithm has input variables obtained by fire experiment with K-type thermocouple and optical smoke sensor. Also triangular membership function is used for inference rules. And the antecedent part of inference rules consists of temperature and smoke density, and the consequent part consists of fire probability. A fire-experiment is conducted with paper, plastic, and n-heptane to simulate actual fire situation. The results show that the intelligent fire detection algorithm suggested in this study can more effectively discriminate signatures between fire and similar fire.