由於傳統的火災偵測器必須充分的接觸目標物（熱或煙霧）才能致動，因此會延遲反應。視覺煙霧檢測比傳統的火災偵測器有更多的優點，例如反應快速、非接觸、大空間監視等。但是大部分的視覺煙霧檢測系統通常容易誤判，發出假警報。
通過分析煙霧的運動特徵，本研究提出區塊化影像結合煙霧運動累積與半遮蔽特性的檢測方式運用在視覺煙霧檢測。由於煙霧通常於燃燒點開始冒出，通過累積方式，找出早期火災的時空視覺特徵，並結合區塊化影像相減的方法，能有效的抑制雜訊的干擾
煙霧有會模糊物體的顏色與部份遮蔽背景的特性，在此提出一種基於分析移動區塊在RGB色彩空間的半遮蔽特性，分析RGB值增加的差異。此方法能有效的呈現出煙霧的半遮蔽特性。
檢測結果表明，結合累積煙霧向上運動與半遮蔽特性的方法，能有效的檢測出煙霧的產生。

Conventional fire detectors must contact these targets（heat or smoke） for being activated and, consequently, respond slowly. Video smoke detection has many advantages over traditional methods, such as fast response and non-contact, large space surveillance, and so on. But most of current methods for video smoke detection systems usually have high false alarms.
By analyzing the characteristics of smoke motion, base on blocking image, a motion accumulation and translucence combined model is proposed for video smoke detection. Because smoke often emerges continually from the place of smoldering, an accumulation model is presented to extract these temporal-spatial visual features of early fire over a time window. The model synthesizes blocking image substrate method can mostly suppress noise.
Smoke can blur colors of objects and partially obscure the object of background. After observing the phenomena, a translucence model is presented base on the RGB color space analysis of the moving block. Analysis the difference of the RGB increases intensity. And the model efficiently represents translucence of the smoke.
Experiments show that the combined accumulation and translucence model is robust and significant for the smoke detection.

[1]. S. Noda, K. Ueda, “Fire Detection in Tunnels sing an Image Processing Method”, Vehicle Navigation and information Systems Conference,pp. 57–62, 1994.
[2]. H. Yamagishi, J.Yamaguchi, “Fire Flame Detection Algorithm Using Color Camera”, Symposium on Micromechatronics and Human Science,1999.
[3]. H. Yamagishi, J. Yamaguchi, “A Contour Fluctuation data Processing Method For Fire Flame Detection Using a Color Camera”, IEEE 26th Annual Conf. on IECON of the Industrial Electronics Society, vol. 2,p22–28, 2000.
[4]. W. Phillips III, M, Shah, N. V. Lobo, “Flame Recognition in Video”,IEEE Workshop on Applications of Computer Vision, 2000
[5]. B. Ugur Toreyin, Y. Dedeoglu, Yigithan, E. Cetin, “Wavelet based real-time smoke detection in video”, 13th European Signal Process Conf, 2005.
[6]. J. Vicente, P. Guillemant, “An Image Processing Technique for Automatically Detecting Forest Fire”, International Journal of Thermal Sciences, 2002.
[7]. J. Fang, J. Jie, Y. H. Yong, Y. M. Zhang, “Early Fire Smoke Movements and Detection in High Large Volume Spaces”, Building and Environment 41 p.1482–1493, 2006.
[8]. T. H. Chen, Y. H. Yin, S. F. Huang, Y. T. Ye, “The Smoke Detection for Early Fire-Alarming System Base on Video Processing”, IEEE Proceedings of the 2006 International Conference on Intelligent Information Hiding and Multimedia Signal Processing, 2006.
[9]. A. J. Kouchinsky, “Determination of Smoke Algorithm Activation For Video Image Detection”, master paper of University of Maryland,2007.
[10]. F. Yuan, “A Fast Accumulative Motion Orientation Model Based on Integral Image for Video Smoke Detection”, Pattern Recognition letters, v.29 n,7, p.925-932, 2008.
[11]. 賴岱佑, “數位影像分析之智慧型監視系統”, 文魁資訊股份有限公司, Ch 4 p.4-9, 2008.
[12]. 連國珍, “數位影像處理”, 儒林圖書公司, 2004.
[13]. 繆紹剛, “數位影像處理”, 台灣培生教育出版股份有限公司, Ch 6,p.301, 2003.
[14]. 惠汝生, “LabVIEW Express 圖控程式應用”, 全華圖書股份有限公司,2004.