Abstract | ||
---|---|---|
Aiming at the difficulty of video smoke detection in complex scene, this paper proposes a method based on DenseNet to identify smoke. By extracting the color features of smoke and the property of upward movement of smoke, the method trains data through dense convolutional neural networks to learn the features between pictures. Experimental results show that the dense convolution neural network model can be effectively applied to real-time detection of smoke events in complex video scenarios.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3299815.3314449 | Proceedings of the 2019 ACM Southeast Conference |
Keywords | Field | DocType |
Color Features, DenseNet, Movement Characteristic, Smoke Detection | Data mining,Pattern recognition,Computer science,Convolutional neural network,Smoke,Artificial intelligence | Conference |
ISBN | Citations | PageRank |
978-1-4503-6251-1 | 0 | 0.34 |
References | Authors | |
0 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
XiQuan Yang | 1 | 0 | 0.34 |
Ying Sun | 2 | 291 | 40.03 |