The Design And Application Of Infrared Ray Fire Surveillance System

Fire Surveillance has been an important task facing the foodstuff system, aswell as leaders at any level is concerned about. National foodstuff annuallylosses caused by fire, theft direct are very alarming, the indirect damagecaused even more difficult to calculate. At the same time in all aspects of oursocial life, the Fire Surveillance is a concern. Responding to the above situationand the problems we developed the Infrared Ray Fire Surveillance System.The thesis first introduced the system hardware as well as thecomposition of each hardware basic design idea, and the summarization offunction of power control panel which I participate in to design.The thesis then introduced the software components of Infrared Ray FireSurveillance System, software components including infrared fire controlsoftware and infrared security surveillance software components. These twoparts can be run independently of each part, but at operating time can mutualcommunications in order to achieve the asynchronous work andsynchronization alarm.In the design and implement of infrared fire control software, I analyzedfour special circumstances of the characteristics of temperature change, andthe filter of the non-flame temperature.First, Jamming signals: The character is that temperature is in thenormal bound, and the duration of jamming signals short but the temperatureis high. The normal filtering method of this signal is scanning around thesudden heat point. If there is no sustained high temperature, the system willcontinue to scan.Second, the impact of seasonal changes: The character is thetemperature changes in same season are small, and the temperature in sameseason is rise or lower with the season change. The temperature in spring andautumn are difference between day and night. The filtering method iscombining the maximum and minimum temperature in weather forecastingwith environment temperature, dynamic storage environment basetemperature.Third, animals or other heat sources: The character is the temperaturelast a long duration, temperature range change less, and the maximumtemperature is below the temperature of fire. The filtering method is judge ifthe temperature of un-normal is in a small range.Fourth, The temperature of fire: The character is the temperature last along duration, the range is not fixed, the change is in a large magnitude, andthe maximum temperature is higher than normal temperature.Through the above analysis, I designed the core modules of the infraredfire control software --Fire Filtering Modules. The modular was design forgive an exactly alarm information and prevent misinformation. On receivinga temperature exceeds a certain value to the environmental temperature, thesoftware control the detector scanning in a small area and analysis if it is afire alarm.In infrared fire control software, I designed the procedure ofasynchronous work and synchronize alarm, the basic principles of this isthrough the initial location and the height of detector, and the vertical andhorizontal steps calculate out the position of the alert. Then according to theposition and the initial position and height of the camera calculate out thevertical and horizontal steps that need to run.In the design and realization of infrared security surveillance software, Idesign a dynamic detection and tracking method which based on backgroundimage processing method, projection method and template matching method,and design a method to save warning video before alarm.The realization advance save of video is that: when the movement waslaunched on the software, system request a disk space, then establish of acycle list, each node of the list stored a frame data, and then create a thread toreceive the frame data into node. When alert occur, system first save the datain the list, and then save the data from the beginning of the frames. List usedhere not only to achieve a single-cycle pre-storage, but also provide thebackground of the background image processing method.The process of dynamic detection and tracking is that: first use thebackground image processing method to decompose out the mobile object,and then use the projection method to identify the location of moving objectsin the current frame, and finally use the template matching method to decidethe direction of movement.The difficulty in background image processing method is how to updatea background. Based on the frame list, select 10 frames from the N frames(ifthe total number of less than 10 frames, then all out), the select method isfrom the beginning of the last frame every N/10 out a frame. Then calculatethe average value of the selected frames, the result is the background image.After construction of background frame, calculate the result image.Projection method is to determine the location of the moving objects inthe image. It change the result image into two value image and then usevertical and horizontal projection to determine the location of moving objectin the image.In the next frame, object will move certain distance in a certain direction,as the interval between two frame images is short, the distance that the objectmove will not long, according to this feature I use the part image searchmethod. That means enlarged the region determined by projection method,and search in the region for identifying the object. Through two or moretimes identify we can judge the direction of object moves.The final of the thesis narrate the customer's opinion to the Infrared RayFire Surveillance System, the problems encountered when install and theimproved method. And the last is the subsequent development of the system.