| Summary: | 碩士 === 中華大學 === 資訊工程學系碩士班 === 100 === Night video surveillance is crucial to construct an all-weather video surveillance system. However, night video surveillance faces several problems: no color information, low brightness, low contrast and low signal to noise ratio (SNR). These problems introduce the false object detections and missing object detections seriously. In this study, we propose a novel night video surveillance method based on the second-order statistics features to overcome the aforementioned problems. First, the block-based foreground detection and dual foregrounds fusion methods are used to extract the candidate regions in which moving objects may exist. Second, by extracting the second-order statistics features in the candidate moving object regions, we may identify the moving objects via the support vector machines (SVM). In addition, identification moving object states based on temporal information are used to tackle the problem of missing detection. Experimental results show that the detection accuracy can approach 93% with the efficiency 25 fps.
|
|---|
