動態特徵法之俯視型行人計數

Feature-Based Approach for Zenithal People Counting

王振宇(Chen Yu Wang)

指導教授 ： 郭忠義

國立臺北科技大學/電資學院/資訊工程系研究所/碩士(2013年)

https://doi.org/10.6841/NTUT.2013.00549

摘要

本論文主要用於特定公共場所中進行人數進出之計算，將攝影機設置於場所出入口正上方，由上往下垂直拍攝，透過這樣的角度下，可以清楚捕捉進出場所之行人頭頂邊緣與紋理，此邊緣與紋理即為本論文所使用之特徵值，以特徵的方式區分場景中行人與非行人之物件，當取得場景中每一行人之所在位置後，使用動態特徵法持續對每一行人進行追蹤並建立軌跡線，當行人與場景中所建立的一條直線產生相交時，根據行人行進的軌跡進行人數的新增與減少，該直線為本論文透過影像技術所建立之計數線。

關鍵字

行人計數 ； 俯視型攝影機 ； 特徵法

並列摘要

This article uses for people counting in the specific public places. The camera is set
on the top of gateway. The zenithal camera can capture the edge and texture of
pedestrian. It is the features in this article. The proposed method distinguishes the
pedestrian from sense by features. The dynamic feature method is used for pedestrian tracking after pedestrian detection. The dynamic feature method establishes the trajectory of pedestrian and updates the feature data whenever. The number of pedestrian has been increase or decrease when the pedestrian crosses the line in the sense. The line is created by image processing. It calls counter line.

並列關鍵字

People Counting ； Zenithal Camera ； Feature-Based Approach

參考文獻

[27] Venkatesh Bala Subburaman, Adrien Descamps, and Cyril Carincotte, “Counting

[21] Alper Yilmaz, Omar Javed, and Mubarak Shah, “Object Tracking: A Survey,”

[20] Qiang Zhu, Shai Avidan, Mei-Chen Yeh, and Kwang-Ting Cheng, “Fast Human

[22] Tai Yu Lai, Jong Yih Kuo, Yong-Yi FanJiang, Shang-Pin Ma, and Yi Han Liao,

[14] K. Gupta, A. V. Kulkarni, “Implementation of an Automated Single Camera