Forest Fire Smoke Recognition Based on Multiple Feature Fusion

2018 2nd International Conference on Artificial Intelligence Applications and Technologies

计算机科学

Lu, Chang^1 ; Lu, Mingqi^2,3 ; Lu, Xiaobo^2,3 ; Cai, Min^2,3 ; Feng, Xiaoqiang^4

School of Software Engineering, University of Science and Technology of China, Hefei, Anhui

230026, China^1

School of Automation, Southeast University, Nanjing

210096, China^2

Key Laboratory of Measurement and Control of CSE, Ministry of Education, Southeast University, Nanjing

210096, China^3

Enbo Technology Company Limited, Nanjing

210022, China^4

关键词: BP neural networks;    BP neutral network;    Cascade classifiers;    Cascaded classifiers;    Fire smoke;    Forest fire smokes;    Forest monitoring;    Multiple feature fusion;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/435/1/012006/pdf
DOI  :  10.1088/1757-899X/435/1/012006

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In order to discover the forest fire at the early stage, the video based fire smoke detection system should be developed. Three static features and three dynamic features are selected to recognize the fire smoke through the analysis of forest monitoring video. To solve the fusion problem of different features, a cascade classifier based on BP neutral network is designed. Each feature vector would be assigned a classifier as the first stage of the cascaded classifier. The output of the first level of the classifier used as the input of the second level of the classifier. The weight for each feature that given by the trainer is more scientific. The result of the experiments show that the cascade classifier has better performance compared to the single BP neural network.