Research Article
Fire Detection Using CNN & SVM
@INPROCEEDINGS{10.4108/eai.7-12-2021.2314594, author={Saranya M and Tamilselvan Arunachalam and Nijanthan N}, title={Fire Detection Using CNN \& SVM}, proceedings={Proceedings of the First International Conference on Combinatorial and Optimization, ICCAP 2021, December 7-8 2021, Chennai, India}, publisher={EAI}, proceedings_a={ICCAP}, year={2021}, month={12}, keywords={features svm learner filter pooling cnn alexnet}, doi={10.4108/eai.7-12-2021.2314594} }- Saranya M
Tamilselvan Arunachalam
Nijanthan N
Year: 2021
Fire Detection Using CNN & SVM
ICCAP
EAI
DOI: 10.4108/eai.7-12-2021.2314594
Abstract
Fire detection in outdoor which can't accomplished by sensor. The fire will endanger property and life of humans. An early caution is very important to prevent the damage that could occur to property and life of human. In indoor, there is possibility of false alarm detection due to change that occurs in the environment. So the best alternative solution to avoid the false alarm is the Image processing based fire detection. Using MATLAB software as a tool in image processing, we can identify whether there is fire or not. This paper proposes a classifying algorithm for identification of the input image as fire or non-fire case based on some statistical features obtained from the images. The extracted features from the input image are provided as input to the SVM to classify them. In addition to SVM, we are using CNN for further improvement because CNN is pre-trained model.