Search

The widespread use of renewable energy resources requires more immediate and effective fire alarms as a preventive measure. The fire is usually weak in the initial stages, which is not conducive to detection and identification. This paper validates a solution to resolve that problem by a flame detection algorithm that is more sensitive to small flames. Based on Yolov3, the parallel convolution structure of Inception is used to obtain multi-size image information. In addition, the receptive field of the convolution kernel is increased with the dilated convolution so that each convolution output contains a range of information to avoid information omission of tiny flames. The model accuracy has improved by introducing a Feature Pyramid Network in the feature extraction stage that has enhanced the feature fusion capability of the model. At the same time, a flame detection database for early fire has been established, which contains more than 30 fire scenarios and is suitable for flame detection under various challenging scenes. Experiments validate the proposed method not only improves the performance of the original algorithm but are also advantageous in comparison with other state-of-the-art object detection networks, and its false positives rate reaches 1.2% in the test set.

Video smoke detection is a promising method for early fire prevention. However, it is still a challenging task for application of video smoke detection in real-world detection systems, as the limitations of smoke image samples for training and lack of efficient detection algorithm. This paper proposes a method based on two state-of-the-art fast detectors, a single-shot multi-box detector, and a multi-scale deep convolutional neural network, for smoke detection using synthetic smoke image samples. The virtual data can automatically offer rich samples with ground truth annotations. However, the learning of smoke representation in the detectors will be restricted by the appearance gap between real and synthetic smoke samples, which will cause a significant performance drop. To train a strong detector with synthetic smoke samples, we incorporate the domain adaptation into the fast detectors. A series of branches with the same structure as the detection branches are integrated into the fast detectors for domain adaptation. We design an adversarial training strategy to optimize the model of the adapted detectors, to learn a domain-invariant representation for smoke detection. The domain discrimination, domain confusion, and detection are combined in the iterative training procedure. The performance of the proposed approach surpasses the original baseline in our experiments.