基於視訊動態背景之運動物體偵測研究

Research of moving object detection in dynamic background for video sequences

蔡政芳(Jeng-Fang Tsai)

指導教授 ： 張元翔

中原大學/工學院/資訊工程研究所/碩士(2013年)

https://doi.org/10.6840/cycu201300369

摘要

視訊監控系統的應用目前已相當普遍，但一般以固定攝影機的靜態背景為基礎，其監控視角較小，容易造成監控死角。本論文提出的系統，針對左、右 Pan 運動之攝影機，可在視訊動態背景中偵測運動物體。本系統採用的方法包含：特徵點偵測、金字塔光流法、光流向量之 K-means 分群等。研究結果顯示，在動態背景且運動物體不超過影像50%的情況下，不僅可偵測出運動的剛性物體，亦適用於非剛性運動的物體，偵測率均達90%以上。總結而言，本系統可增加攝影機視角範圍，大幅降低視訊監控的死角。

關鍵字

影像金字塔 ； 光流 ； 特徵點

並列摘要

Applications of video surveillance systems are commonly seen at present. These systems are commonly based on a fixed camera that yields static background and a narrow angle-of-view, therefore are subject to surveillance limitations. This thesis proposed a system in an attempt to detect moving objects with dynamic background in digital videos, using a moving camera which may pan left or right. This system adopts the methods including: feature point detection, pyramid optical flows, K-means clustering, etc. Our results demonstrate that our system is able to detect moving rigid objects as well as non-rigid objects, given the situation if the moving object with dynamic background doesn’t occupy over 50% of the image area. The overall detection rate has achieved over 90%. In summary, our system is able to increase the angle-of-view, thus reducing the surveillance limitations (blind spot).

並列關鍵字

image pyramid ； K-means ； feature point ； optical flow

參考文獻

[2] A. J. Lipton, H. Fujiyoshi, and R. S. Patil, “Moving target classification and tracking from real-time video,” Proc. of the IEEE Workshopon Applications of Computer Vision, pp. 8-14, 1998.

[3] B. K. P. Horn and B. G. Schunck, “Deterimining optical flow,” Artificial Intelligence, Vol. 17, pp. 185-203, 1981.

[4] C. Kim and J. N. Hwang, “Fast and automatic video object segmentation and tracking for content-based applications,” IEEE Transactions Circuits and Systems for Video Technology, Vol. 12, pp. 122-129, 2002.

[5] A. G. Nguyen and J. N. Hwang, “Scene context dependent key frame selection in streaming,” in Proc of the 22nd International Conference on Distributed Computing Systems Workshops, pp. 208-213, 2002.

[7] K. Y. Chase, W. P. Dong, and C. S. Jeong, “SUSAN window based cost calculation for fast stereo matching [J] ,“ Lecture Notes in Computer Science, Vol. 38(2), pp. 947-952, 2005.