Combining Neural Networks and a Color Classifier for Fire Detection

Deep learning methods have solved several problems in the computer vision area, mainly for image classification. The use of these methods for fire detection can bring great improvements to security systems and prevent many losses. Effective fire detection systems help fire situations to have lesser consequences than they could have by signaling as quickly as possible the experts. This paper proposes a combination of modern image classification models and a color classifier to detect and localize the fire. The CoAtNet-4 architecture is used to detect and localize fire in still images, while the color classifier refines the mask obtained. State-of-the-art works show a high number of parameters, imbalanced results and lower true positive rates (TPR), while our results show high TPR, great accuracy and a balanced classification.