ATT Squeeze U-Net A Lightweight Network for Forest Fire Detection and Recognition

Aim:

To efficient CNN based system for fire detection in videos captured in uncertain surveillance scenarios

Synopsis:

          Vision based fire detection framework has lately picked up popularity when contrasted with customary fire recognition framework dependent on sensors. The need of video perception at private, Modern, business regions and woods areas has expanded the use of vision based fire acknowledgment system Recently lots of fire related accidents has occurred due to improper Surveillance or unable to cover those uncertain regions like restricted areas in forest or any factory buildings. In order to overcome such accidents , we propose a new method using Convolutional neural networks (CNN). To solve these problems, a more advanced fire detection scheme proposing the use of CNN technology. instead of feature description has attracted more and more attention. In this article, we propose an efficient neural network architecture for forest fire detection and recognition based on CNN.

Proposed System:

           We propose an efficient CNN based system for fire detection in videos captured in uncertain surveillance scenarios. Our approach uses light-weight deep neural networks with no dense fully connected layers, making it computationally inexpensive. Once fire detected the information will pass through the firebase. Firebase is used as cloud database and to send notification that can be received in android smartphone.