Military Application of Intelligent Monitoring System

• Main Authors: Chen,Kuo-wen, 陳國文
• Other Authors: Lin, Jenn-Huei
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/49198197061174974523

碩士 === 國防大學理工學院 === 電子工程碩士班 === 99 === Background subtraction is a fundamental approach for image sequences in the image processing, The various methods differ in type of background model and technology used to update the model proposed and implemented continuously. Therefore, a detection technique, as to challenges the fast and effective real-time system, is basic and the most important issue. In particular, the coexistence of the environment in the face of illumination and shadow is especially true. For long real-time monitoring system between the image clarity and computing efficiency there has been a problem of trade-offs. In this thesis, taking into account the real-time and accuracy of the real-time monitoring system at the same time. The first step used the improvement background subtraction method, then sampling combined with the correlation operation between pixels. Identify with making a reasonable threshold to extract rapidly invasion of images from the continuous monitoring system. Then, using HSV color space threshold set to remove the shadow. Perform the image processing and recognition of work-related. The second step combined webserver database system to record the data and in-depth analysis. It provided remote alarm and detailed identification. Experimental results demonstrate that the proposed method can detect quickly the intrusion of real-time system. Invasion information were also recorded in the web database system to provide remote query.Therefore, it’s very helpful for protection of military facilities(i.e.oil depot,gun room,ammunition depot ,Combat situation room etc.).