| Titre : | Detection of smoke and fire using Convolutional NeuralNetworks And the real-time Yolo detector |
| Auteurs : | Abd elrezak Chantouf Larbi, Auteur ; acine Abd El Hamid Benbahi Taha, Y, Auteur ; Rochdi Bachir Bouiadjra, Directeur de thèse |
| Type de document : | texte manuscrit |
| Editeur : | Université mustapha stambouli de Mascara:Faculté des sciences exactes, 2023 |
| ISBN/ISSN/EAN : | SE02338T |
| Format : | 75P. / couv. ill. / 29cm. |
| Accompagnement : | disque optique numérique (CD-ROM) |
| Langues: | Anglais |
| Résumé : |
In conclusion, our study has demonstrated the effectiveness of using YOLO (You Only Look Once) and CNN (Convolutional Neural Network) for forest fire detection. Through these state-of-the-art computer vision techniques, we were able to develop a system capable of analyzing real-time images from various sources, such as drones, surveillance cameras, or satellites, to quickly identify forest fires and alert the relevant authorities. The YOLO approach proved to be particularly suitable for our problem, as it allows for the real-time and highly accurate detection of forest fires. By using a CNN to extract relevant features from the images, we trained our model to recognize different components of a forest fire, such as flames, smoke, and areas of intense heat. Our experiments have shown that our forest fire detection system based on YOLO and CNN can achieve high performance, with a precise detection rate and a low false positive rate. This suggests that our approach can be a valuable tool for firefighting services and natural resource managers, enabling them to respond quickly to forest fires and minimize environmental and human damage. However, it is worth noting that our system is not without limitations. For example, it may encounter difficulties in extreme weather conditions, such as thick smoke or low visibility due to dense fog. Additionally, our model is specific to forest fires and cannot detect other types of fires. To move forward, it would be interesting to further research in the field of forest fire detection by leveraging larger and more diverse datasets, as well as refining the architectures of the neural networks used. Furthermore, the integration of advanced techniques such as transfer learning or reinforcement learning could further improve the system’s performance. In conclusion, our thesis has demonstrated that the use of YOLO and CNN for forest fire detection is a promising approach. Through this technology, we can hope for a faster and more effective response to forest fires, thereby contributing to the preservation of our precious forest ecosystems and the protection of human lives |
Exemplaires (1)
| Code-barres | Cote | Support | Localisation | Section | Disponibilité |
|---|---|---|---|---|---|
| SE02338T | INF924 | Livre audio | Bibliothèque des Sciences Exactes | 7-Mémoires Master | Consultation sur place Exclu du prêt |
Aucun avis, veuillez vous identifier pour ajouter le vôtre !
Accueil
