Fog Computing and Deep CNN Based Efficient Approach to Early Forest Fire Detection with Unmanned Aerial Vehicles
| Autor: | Kethavath Srinivas, Mohit Dua |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Deep cnn
Disaster monitoring business.industry Computer science Fire detection Real-time computing ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Response time 020206 networking & telecommunications Cloud computing 02 engineering and technology Convolutional neural network Fog computing 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Internet of Things business |
| Zdroj: | Inventive Computation Technologies ISBN: 9783030338459 |
| Popis: | Fog computing assits the development of distributed real-time systems. This offers solutions to a quicker response systems for developing disaster monitoring, prevention and detection models into existence. This paper proposes the integration of Fog computing and Convolutional Neural Networks (CNN) with Unmanned Aerial Vehicles (UAV) to detect the forest fire at an early stage. A highly efficient CNN model has been used for fire image recognition due to its proven ability for such recognition tasks. By using AlexNet and other architectures in the proposed model, image recognition tasks have become more capable, to an extent that a pre-trained model has an ability equal to a primate. Using these architectures, we trained our model and deployed the same on a Fog device, which has resulted in achieving higher accuracy and response time. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|