Comparison of two deep learning methods for detecting fire hotspots

Every high-rise building must meet construction requirements, i.e. it have good safety to prevent unexpected events such as fire incident. To avoid the occurrence of a bigger fire, surveillance using closed circuit television (CCTV) videos is necessary. However, impossible for security forces monitor full day. One methods that can be used help deep learning method. In this study, we use two detect hotspots, you only look once (YOLO) method and faster region-based convolutional neural network (faster R-CNN) The first stage, collected 100 image data (70 training 30 test data). next stage model which aims make recognize fire. Later, calculate precision, recall, accuracy, F1 score measure performance model. If close 1, then balance optimal. our experiment results, found YOLO has precision 100%, recall 54.54%, accuracy 66.67%, 0.70583667. While R-CNN 87.5%, 95.45%, 86.67%, 0.913022.