Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs
The technologies underlying fire and smoke detection systems play a crucial role in ensuring and delivering optimal performance in modern surveillance environments. In fact, fire can cause significant damage to lives and properties. Considering that the majority of cities have already installed came...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-11-01
|
| Series: | Atmosphere |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4433/11/11/1241 |
| _version_ | 1797547553585102848 |
|---|---|
| author | Yakhyokhuja Valikhujaev Akmalbek Abdusalomov Young Im Cho |
| author_facet | Yakhyokhuja Valikhujaev Akmalbek Abdusalomov Young Im Cho |
| author_sort | Yakhyokhuja Valikhujaev |
| collection | DOAJ |
| description | The technologies underlying fire and smoke detection systems play a crucial role in ensuring and delivering optimal performance in modern surveillance environments. In fact, fire can cause significant damage to lives and properties. Considering that the majority of cities have already installed camera-monitoring systems, this encouraged us to take advantage of the availability of these systems to develop cost-effective vision detection methods. However, this is a complex vision detection task from the perspective of deformations, unusual camera angles and viewpoints, and seasonal changes. To overcome these limitations, we propose a new method based on a deep learning approach, which uses a convolutional neural network that employs dilated convolutions. We evaluated our method by training and testing it on our custom-built dataset, which consists of images of fire and smoke that we collected from the internet and labeled manually. The performance of our method was compared with that of methods based on well-known state-of-the-art architectures. Our experimental results indicate that the classification performance and complexity of our method are superior. In addition, our method is designed to be well generalized for unseen data, which offers effective generalization and reduces the number of false alarms. |
| first_indexed | 2024-03-10T14:45:42Z |
| format | Article |
| id | doaj.art-43a523d761ba4fdaafb5c4c135e3386d |
| institution | Directory Open Access Journal |
| issn | 2073-4433 |
| language | English |
| last_indexed | 2024-03-10T14:45:42Z |
| publishDate | 2020-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Atmosphere |
| spelling | doaj.art-43a523d761ba4fdaafb5c4c135e3386d2023-11-20T21:22:06ZengMDPI AGAtmosphere2073-44332020-11-011111124110.3390/atmos11111241Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNsYakhyokhuja Valikhujaev0Akmalbek Abdusalomov1Young Im Cho2Department of IT Convergence Engineering, Gachon University, Sujeong-Gu, Seongnam-Si, Gyeonggi-Do 461-701, KoreaDepartment of IT Convergence Engineering, Gachon University, Sujeong-Gu, Seongnam-Si, Gyeonggi-Do 461-701, KoreaDepartment of Computer Engineering, Gachon University, Sujeong-Gu, Seongnam-Si, Gyeonggi-Do 461-701, KoreaThe technologies underlying fire and smoke detection systems play a crucial role in ensuring and delivering optimal performance in modern surveillance environments. In fact, fire can cause significant damage to lives and properties. Considering that the majority of cities have already installed camera-monitoring systems, this encouraged us to take advantage of the availability of these systems to develop cost-effective vision detection methods. However, this is a complex vision detection task from the perspective of deformations, unusual camera angles and viewpoints, and seasonal changes. To overcome these limitations, we propose a new method based on a deep learning approach, which uses a convolutional neural network that employs dilated convolutions. We evaluated our method by training and testing it on our custom-built dataset, which consists of images of fire and smoke that we collected from the internet and labeled manually. The performance of our method was compared with that of methods based on well-known state-of-the-art architectures. Our experimental results indicate that the classification performance and complexity of our method are superior. In addition, our method is designed to be well generalized for unseen data, which offers effective generalization and reduces the number of false alarms.https://www.mdpi.com/2073-4433/11/11/1241fire detectionsmoke detectiondeep learningdilated convolutionclassification |
| spellingShingle | Yakhyokhuja Valikhujaev Akmalbek Abdusalomov Young Im Cho Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs Atmosphere fire detection smoke detection deep learning dilated convolution classification |
| title | Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs |
| title_full | Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs |
| title_fullStr | Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs |
| title_full_unstemmed | Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs |
| title_short | Automatic Fire and Smoke Detection Method for Surveillance Systems Based on Dilated CNNs |
| title_sort | automatic fire and smoke detection method for surveillance systems based on dilated cnns |
| topic | fire detection smoke detection deep learning dilated convolution classification |
| url | https://www.mdpi.com/2073-4433/11/11/1241 |
| work_keys_str_mv | AT yakhyokhujavalikhujaev automaticfireandsmokedetectionmethodforsurveillancesystemsbasedondilatedcnns AT akmalbekabdusalomov automaticfireandsmokedetectionmethodforsurveillancesystemsbasedondilatedcnns AT youngimcho automaticfireandsmokedetectionmethodforsurveillancesystemsbasedondilatedcnns |