A Study on the Real-time Fire Classification Algorithm Based on Stacked LSTM

Several researchers have been drawn to the development of fire detector in recent years, protect people and property from catastrophic disaster fire. However, studies related monitoring are affected by some unique characteristics sensor signals, such as time dependence complexity signal pattern based on variety types,. In this study, a new deep learning-based approach that accurately classifies various types situations real-time using data obtained multidimensional channel signals was proposed. The contribution study is develop stacked-LSTM model considers time-series sensing framework for identification improving existing detectors.