Early Detection of Forest Fire using Deep Learning

Abstract

Forest fires have become a serious threat to mankind. Besides providing shelter and protection to a large number of living beings, they have been a major source of food, wood, and a great supply of other products. Since ancient times forests have played an important role in social, economic, and religious activities and have enriched human life in a variety of ways both material and psychological. To protect our nature from these rapidly rising forest fires, we need to be cautious enough of every decision we take which could lead to a disastrous end, once and for all. So for the early detection of forest fires, we propose an image recognition method based on Convolutional Neural Networks (CNN). We have fine-tuned the Resnet50 architecture and added a few convolutional layers with ReLu as the activation functions, and a binary classification output layer which showed a huge impact on the training and test results when compared to the other SOTA methods like VGG16 AND DenseNet121. We achieved a training set accuracy of 92.27% and 89.57% test accuracy with a stochastic gradient descent optimizer and we have avoided the underfit/overfitting on the model with the help of the Stochastic Gradient Descent (SGD) algorithm.