人流追蹤與計數方法之研究

Study of Pedestrian Tracking and Counting Method

陳忠宏(Chung-Hung Chen)

指導教授 ： 黃有評

國立臺北科技大學/電資學院/電機工程系研究所/碩士(2014年)

摘要

近年來隨著資訊科技的進步，使用電腦視覺的應用也越來越廣泛。而人數統計在出入口及公共場所中對於安全監控有相當高的參考價值，過去通常都是以人工或傳統方式來進行人數控管，但是這些方法必需要花費較高的人力、物力及成本。因此，本研究利用電腦視覺處理方式結合智慧型安全監控系統對監控區域內進行行人檢測、人流追蹤及人數統計。首先透過攝影機來擷取行人進出的影像，結合影像處理的技巧來分割視訊物件，再針對各個物件使用Haar特徵分類器進行行人檢測辨識，之後藉由CAMSHIFT追蹤演算法進行人流追蹤，並且使用Kalman濾波器來預估追蹤物件的軌跡作為輔助以提升追蹤的效能，最後透過虛擬線判定的方式計算離開及進入區域的人數，取得人流數量資料。經由實驗證實了此方法的可行性，而達到節省人力資源、降低設備成本的需求。

關鍵字

人流計數 ； CAMSHIFT追蹤 ； Haar特徵分類器 ； Kalman濾波器

並列摘要

With advances in information technology in recent years, the applications of computer vision are increasingly widespread. The people counting in the entrances to public places have very high reference values to security monitoring. Conventional methods usually used manual or traditional way to count the number of people, but they required a lot of manpower, material resources and costs. Therefore, this study combines computer vision approach and intelligent security surveillance system to detect, track and count people in the monitored areas. First, we use the camera to capture images of pedestrians crossing the region and apply integrated image processing techniques to split objects. Then, we use Haar feature classifier to identify each object and track them by CAMSHIFT algorithms. In addition, the Kalman filter is adopted to estimate the trajectory of the object as an auxiliary track to improve tracking performance. Finally, the number of pedestrians is counted in the region to obtain quantitative information by virtual lines. The experimental results demonstrate the feasibility of the proposed method. The proposed work has the merits of savings in manpower resources and reduction in equipment costs.

並列關鍵字

people counting ； CAMSHIFT tracking ； Haar feature classifier ； Kalman filter

參考文獻

[8] Z. Lin, L.S. Davis, D. Doermann and D. DeMenthon, “Hierarchical part-template matching for human detection and segmentation,” in Proc. of IEEE 11th International Conference on Computer Vision, Rio de Janeiro, Portugal, pp.1-8, Oct. 2007.

[3] O. Masoud and N.P. Papanikolopoulos, “A novel method for tracking and counting pedestrians in real-time using a single camera,” IEEE Trans. on Vehicular Technology, vol. 50, no. 5, pp.1267-1278, Sep. 2001.

[9] V. Rabaud and S. Belongie, “Counting crowded moving objects,” in Proc. of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, New York, NY, USA, vol. 1, pp.705-711, Jun. 2006.

[10] L. Xu, J. Jia, and Y. Matsushita, “Motion detail preserving optical flow estimation,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 34, no. 9, pp.1744-1757, Sep. 2012.

[11] X.J. Song, L.D. Seneviratne and K. Althoefer, “A Kalman filter-integrated optical flow method for velocity sensing of mobile robots,” IEEE/ASME Trans. on Mechatronics, vol. 16, no. 3, pp.551-563, Jun. 2011.