發(fā)布時間：2018-12-31 11:52

【摘要】：火災(zāi)是一種具有較強突發(fā)性和破壞性的自然災(zāi)害。尤其在大空間建筑中，特殊的建筑結(jié)構(gòu)與建筑用途使火災(zāi)發(fā)生的幾率增大，蔓延的速度加快，危害的程度加深，預(yù)防與撲救的難度加大，使得消防任務(wù)極其艱巨。圖像型火災(zāi)探測與智能消防炮定點撲救系統(tǒng)是一項針對大空間建筑防滅火而研發(fā)的消防利器，系統(tǒng)可以實時對現(xiàn)場進行火災(zāi)監(jiān)控，對火災(zāi)進行報警并定位，同時指揮消防炮進行滅火撲救工作。 本文以前期火災(zāi)探測系統(tǒng)已經(jīng)準(zhǔn)確發(fā)現(xiàn)火災(zāi)源點為前提，重點研究了大空間建筑中火災(zāi)的圖像定位技術(shù)。主要依據(jù)計算機視覺、數(shù)字圖像處理、光學(xué)測量等技術(shù)原理，實現(xiàn)對火災(zāi)源點的空間初定位與精定位，最終得出較為精確的火災(zāi)源點在大空間建筑中的三維坐標(biāo)，提供給消防炮作為滅火依據(jù)。本文定位系統(tǒng)的流程主要包括：在已經(jīng)探測出的視頻火災(zāi)圖像上進行特征提取、特征匹配、空間位置的計算等過程。本文吸取靶場光電測量的高精度優(yōu)點，并使之適合大空間建筑的遠程測量，并增加了光軸重合的定位方法共同組成遠程測量系統(tǒng)，在近程測量上保留了平行雙目定位技術(shù)進行補充。在此基礎(chǔ)上得到初定位數(shù)據(jù)后，指揮消防炮進行粗瞄準(zhǔn)，，對初定位結(jié)果進行光束指示，結(jié)合已知觀察信息二次對攝像機進行標(biāo)定，并提取光束光斑圖像與火災(zāi)源點圖像進行比對。依照火源位置，對遠程數(shù)據(jù)、近程數(shù)據(jù)與比對數(shù)據(jù)進行多傳感器數(shù)據(jù)融合，修正火災(zāi)位置信息，給出精定位結(jié)果。實驗證明，算法在大空間建筑火災(zāi)定位的應(yīng)用中較好的滿足了精度要求，提高了自動滅火的效率。
[Abstract]:Fire is a kind of natural disaster with strong sudden and destructive. Especially in the large space building, the special building structure and the building use increase the probability of fire, accelerate the speed of spread, deepen the degree of harm, increase the difficulty of prevention and rescue, make the task of fire fighting extremely arduous. The image fire detection and intelligent fire fighting system is a fire fighting weapon developed for large space building. The system can monitor the fire on the spot, alarm and locate the fire in real time. At the same time command the fire gun fire suppression work. Based on the premise that the fire detection system has found the fire source accurately, this paper focuses on the image location technology of fire in large space building. Based on the principles of computer vision, digital image processing and optical measurement, the initial location and precise location of the fire source are realized, and the 3D coordinates of the fire source in the large space building are obtained. Provide fire fighting gun as fire extinguishing basis. In this paper, the process of location system mainly includes: feature extraction, feature matching, spatial location calculation and so on. In this paper, the advantages of high accuracy in range photoelectric measurement are absorbed, and it is suitable for the remote measurement of large space buildings, and the optical axis coincidence positioning method is added together to form a remote measurement system. Parallel binocular positioning technique is reserved for short-range measurement. On this basis, the initial positioning data are obtained, and the fire fighting gun is directed at the initial position, the initial positioning result is indicated by the beam, and the camera is calibrated twice with the known observation information. The beam spot image is compared with the fire source image. According to the location of the fire source, the remote data, the short-range data and the comparison data are fused to the multi-sensor data, the fire location information is corrected, and the precise location results are given. Experimental results show that the algorithm can meet the requirements of accuracy and improve the efficiency of automatic fire extinguishing in large space building fire location.
【學(xué)位授予單位】：西安建筑科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2013
【分類號】：TU892;TP391.41

【參考文獻】

相關(guān)期刊論文 前10條

1 湯正華,王殊,陳濤;多傳感器/多判據(jù)探測器在火災(zāi)探測中的應(yīng)用[J];傳感器技術(shù);2001年03期

2 郭鍵,王汝琳,李明;火災(zāi)探測技術(shù)的現(xiàn)狀及發(fā)展方向[J];遼寧工程技術(shù)大學(xué)學(xué)報;2004年02期

3 陳濤,袁宏永,蘇國鋒,劉炳海;基于CCD圖像的火災(zāi)空間定位方法[J];光學(xué)技術(shù);2003年03期

4 郭黎;廖宇;陳為龍;;基于改進SIFT的視頻超分辨率重建快速配準(zhǔn)算法研究[J];湖北民族學(xué)院學(xué)報(自然科學(xué)版);2010年02期

5 孟浩;程康;;基于SIFT特征點的雙目視覺定位[J];哈爾濱工程大學(xué)學(xué)報;2009年06期

6 王偉,吳成柯;估計基礎(chǔ)矩陣的六點綜合算法[J];中國科學(xué)E輯:技術(shù)科學(xué);1997年02期

7 朱艷,劉方,蒲清平;大空間建筑消防安全評估[J];重慶建筑大學(xué)學(xué)報;2005年02期

8 羅曉暉;杜召彬;;MSA火災(zāi)報警系統(tǒng)中圖像型火源定位研究[J];計算機工程與應(yīng)用;2009年22期

9 許維勝,田長征,方盛明;基于圖像視覺特征的火災(zāi)自動識別[J];計算機工程;2003年18期

10 吳福朝,胡占義;攝像機自標(biāo)定的線性理論與算法[J];計算機學(xué)報;2001年11期

相關(guān)博士學(xué)位論文 前1條

1 閔永林;大空間智能消防水炮滅火系統(tǒng)研究[D];上海大學(xué);2010年

本文編號：2396534