Anomaly Foreground Detection through Background Learning in Video Surveillance视频监控中背景学习的异常前景检测.pdfVIP

Anomaly Foreground Detection through Background Learning in Video Surveillance 1 Cheng-Yuan Tang, Yi-Leh Wu, Shih-Pin Chao, Wen-Chao Chen, and Pan-Lan Chen Abstract. We present a new set of rapid detection of background subtraction algo- rithms using codebooks to established Background Model (BG Model) and the concept of Color Model originally proposed by [6]. The proposed methods do not require prior learning, as in [6], and can create an instant BG Model detection and training with instant learning mechanism. Our proposed methods can also turn the latter coming but stationary foreground objects gradually as background, which are more adaptive to the actual environments. The proposed methods can also use instant learning to absorb sudden camera movements caused by the environments. We show that the proposed methods are effective and efficient in video surveil- lance applications. Keywords: foreground detection, codebook, background model. 1 Introduction In the study of visual surveillance, effective background detection has been a very important issue. A common approach to capture moving foreground is to use back-ground subtraction with the pure background and the current image. Tradi- tional methods require a precise static background and thus cannot be used prop- erly in a dynamic environment where the background changes dramatically. Some re-searchers propose to establish a background model by a series of statistical color images, by the time sequence of each color pixel to establish a single distribution model [1-2,7-8]. However, the background model of a single color Cheng-Yuan Tang and Pan-Lan Chen Department of Information Management, Huafan Uniersity, Taiwan Yi-Leh Wu Department of Computer Science and Information Engineering, National Taiwan University of S