Feature Extraction And Analysis System For Flame Radiation Data

In war, the oil box of a tank or armored vehicle would be fired and exploded when it was attacked by cannonballs. The explosion can lead to a more serious injury, even fatal injury. In order to avoid the injury, armored vehicles should be equipped with high-speed fire fighting equipments. The most important part of the facility is a flame detector. The capability of flame detector plays a crucial role in detecting and putting out the fire. In order to developing the high-sensitive and high-accurate flame detector, the flame radiation data should be analyzed and the character parameter should be obtained.In the original foundation of a set of hardware equipment used for collect flame radiation data, I designed a software used for analyzing the spectrum data in this article,which are used to detect the standard flame in the lab. The foundation data of flame radiation can be received by using this equipment, which is able to analyze the flame radiation data. All these can make us acquainted with the basic characters of the flame radiation data and lay a foundation of further developing high-powered optic spectrum-type fire forecast detector. In the article, a spectral library is built up using an Object Oriented programming language and relational database managing software. The radiation spectrum data can be input and output expediently. Moreover, the analysis of radiation spectrum data has been completed, including picking up character parameters of the spectrum, derivative spectral analysis and so on. In addition, the software uses Mahalanobis Distance to analyze whether there is fire in the target and the comparability analyze of the flame targets. All these can acquaint ourselves with the characteristics of the flame spectrum and lay a foundation for further developing high-sensitive spectrum-type fire forecast system.