用球型攝影機做電腦視覺的物件追蹤

Computer Vision Target Tracking Using a Fast Dome Camera

陳昰亘(Shin-Ken Chen)

指導教授 ： 林啟芳

元智大學/資訊學院/資訊工程學系/碩士(2004年)

摘要

本論文的主要目的是在攝影監控系統中提出物件追蹤的方法，利用可以旋轉的球型攝影機去追蹤與放大被鎖定的物體，以便獲得被鎖定物體的較清晰影像。方法內容主要是利用一部個人電腦來擷取球型攝影機所拍攝到的畫面，分析畫面的色彩並透過兩次的連續可適性質量中心偏移（CAMSHIFT）演算法找出鎖定物體的位置及大小。第一次連續可適性質量中心偏移演算法是用來修正預測的形狀遮罩位置，第二次則是計算追蹤物體的偏移量進而找出物體在畫面中的位置及大小。計算連續影像中物體的位置及大小變化可以得知物體的移動方向與速度等資訊，再根據這些資訊來控制球型攝影機，讓攝影機鏡頭可以隨著被鎖定物體移動。

關鍵字

目標追蹤 ； 色彩物體擷取 ； 球型攝影機

並列摘要

This paper presents a method to track an object in a surveillance system using a dome camera, having the abilities of pan, tilt and zooming. The main purpose of tracking is to track the trajectories of the moving target, and obtain better images of the target. The content of the method includes capturing images, analyzing the color of images, and performing a tracking algorithm twice to find the position of the target. The first time of the tracking algorithm is to correct the predicted position of the shape mask, and the second time is to find the position and the size of the target. Positions of the target in continuous frame rate will give information about the trajectories that the target moves. Finally, according to the obtained information, we could control the dome camera so that it can follow the moving target.

並列關鍵字

Target Tracking ； Color Segmentation ； Dome Camera

參考文獻

[3] Shao-Yi Chien, Shyh-Yih Ma, and Liang-Gee Chen, “Efficient moving object segmentation algorithm using background registration technique,” IEEE Transactions on Circuits and Systems for Video Technology, Vol. 12, Issue 7, pp. 577-586, July 2002.

[4] Changick Kim and Jeng-Neng Hwang, “Fast and automatic video object segmentation and tracking for content-base applications,” IEEE Transactions on Circuits and Systems for Video Technology, Vol. 12, Issue 2, pp. 122-129, Feb. 2002.

[5] A. Shio and J. Sklansky, “Segmentation of people in motion,” Proceedings of IEEE Workshop on Visual Motion, pp. 325-332, 1991.

[6] 許詠泰，“應用於監視系統之強健式整合性的即時移動物體偵測與追蹤演算法”，元智大學，機械工程研究所，碩士論文，民國92年。

[7] A. M. Baumberg and D. C. Hogg, “An efficient method for contour tracking using active shape models,” Proceedings of IEEE workshop on Motion of Non-Rigid and Articulated Objects, pp. 194-199, 1994.