Conflagration Recognition Using Convolution Neural Networks With Warning System

Document Type : Primary Research paper


1 Assistant Professor, Department of CSE, Bannari Amman Institute of Technology, Sathyamangalam,

2 Research Scholar, Department of CSE, KPR Institute of Technology, Coimbatore,

3 Professor, Department of IT, Bannari Amman Institute of Technology, Sathyamangalam


The main objective of the project is to enhance the existing system for fire
prediction and protection systems. Deep Learning approaches have shown the ability to
provide better results for the prediction of wildfires. Previous systems are more
complicated and sort of complete black box in image analysis. Faster recognition and
passing the data to the concerned wildfire authorities in more important. The warning
system needs to be automated and should send the information at a set of intervals for
proper image analysis. Hence, an automatic warning system to the system is suggested to
avoid late actions taken against the wildfire and its damages. Also different types of image
analysis for categorizing the fires occurred is proposed. DWT is one of the image analysis
methods that has been implemented for better result. This will help any people with less
knowledge about the wildfire to understand the fire nature and enables them to take
appropriate actions. The neural network model in the Restnet50 was faster in training with
large datasets compared to other neural network models.


Volume 12, Issue 3 - Serial Number 3
ICMMNT-2021 International Virtual Conference on Materials, Manufacturing and Nanotechnology, 30th June, 2021.
June 2021
Pages 1294-1300