Design Of Flame Image Acquisition And Analysis System Based On Infrared Thermal Image

Fire has been a cause of great damage to life and property safety all the time.To effectively avoid any huge loss caused by fire,the firefighting undertaking rises in response.In the wake of constant technical progress in recent years,firefighting is technically evolving and increasing emerging technologies are applied to firefighting.Among links of firefighting,flame detection is paramount in fire warning.At the moment,various flame detection methods are available.The detection method by the infrared thermal imaging instrument stands out in particular on account of its outstanding performance in temperature detection,i.e.,significantly good effects of detecting anything in high temperature,which has been used in a broad range in firefighting.Based on infrared thermography technology and aimed at flame detection,a flame image detection algorithm based on infrared thermography technology is put forward in this thesis and applied to scenarios in request for close-range and high-precision monitoring in particular.On the ground of this algorithm,a flame detection system software package based on infrared thermography is developed.Relevant details are set forth as follows:(1)A flame image detection algorithm based on infrared thermography technology is proposed.This method has sound robustness to high-temperature disturbance.First of all,an infrared thermal imager is used to acquire infrared radiation data in a scenario of detection and relevant data are quickly put into temperature approximation and conversion to arrive at temperature approximation image.Subsequently,key areas of the temperature approximation image and flame-related image features in the key areas are extracted.Further to relevant multi-frame feature data,flame determination is done to the feature data in the support vector machine(SVM)algorithm.Last but not least,the algorithm proposed is used to distinguish fuel flame and plastic flame from such disturbance terms as the tungsten lamp,heating lamp,hair dryer,and high-temperature iron bowl.Final test results show that the algorithm in this thesis can accurately detect 99.9% of flame.(2)A flame image acquisition and analysis software system based on infrared thermography is also designed and implemented in this thesis.It is a flame detection software system that applies to Windows systems,which results from the preceding flame image detection algorithm based on infrared thermography.In design,this software system comprises a software architecture,management control module,infrared thermal imager communication module,flame detection algorithm module,infrared image display module,maximum temperature warning module,scenario recording module and alarm module.The system implements such primary functions as acquiring image data by infrared thermal imager and monitoring fire in a scenario.In the event of fire,the system can give a fire alarm in time and record relevant information;besides,the system can also record maximum temperature information related to the scenario under surveillance on a daily basis for reference of fire prevention in the scenario.