視訊監視中有關攝影機異常偵測與分類

數位攝影機已廣泛使用在居家安全、路口監視、社區防護等場所，成為醫療照護與嚇阻犯罪的重要利器。隨著監視器數量與日俱增，監視人員需同時觀看大量的監視畫面，因而容易忽略部份的拍攝內容。為了維護系統的正常運作，監視系統應可自動偵測異常情況，以避免人為或其他因素的破壞而失去該有的監視功能。此一研究首先探討可能的原因，並提出有關監視攝影機之破壞偵測方法。除了對常見的遮蔽、失焦、位移等三類異常狀況作即時偵測外，尚能區分遭受的破壞類型，作為狀況排除的主要依據。同時考量室內與戶外多變的環境，以實測方法驗證系統的可靠性。

關鍵字

遮蔽偵測；失焦偵測；位移偵測；攝影機破壞偵測

並列摘要

Digital cameras had been used widely in applications of home security, road monitoring, community guard, and become an important part of health care and deterring crime.Because of increaseddigital cameras day by day, monitoring staffs need to watch numerous of screens at the same time so that it’s easy to losesome of details.To maintain the system to work normally, monitoring system should be able to detect the abnormal situation automatically, and avoid human factors or other reasons to lose the monitoring function. This research firstly investigates the possible reasons and proposes the way to detect the unusual monitoring camera.Besides the obstruction, fuzzy and displacement can be detected immediately, this system could also classify different types of damage as the main basis for status exclusion.Finally,some indoor and outdoor environments are given to test empirically for reliability evaluation.

並列關鍵字

obstruction detection ； fuzzy detection ； displacement detection ； camera tamper detection

參考文獻

[2]詹子厚，日夜視訊之人臉辨識，國立交通大學電控工程研究所碩士論文，2014。

[3]S. Agarwal and D. Roth, “Learning a sparse representation for object detection,” ECCV’02, pp. 113-130, 2002.

[8]E. Ribnick, S. Atev, O. Masoud, N. Papanikolopoulos and R. Voyles, “Real-time detection of camera tampering,”IEEE International Conference on Video and Signal based Surveillance, November 2006.

[9]Anil Aksay,AlptekinTemizel and A. Enis Cetin, “Camera tamper detection using wavelet analysis for video surveillance,”IEEE International Conference on Advanced Video and Signal based Surveillance, 2007.

[11]T. Kohonen, “Learning vector quantization,” Neural Networks, vol. 1, pp.3-16, 1988.

國際替代計量

視訊監視中有關攝影機異常偵測與分類

全文下載

主題瀏覽