|Table of Contents|

A method of real-time traffic background updating based on block-counting(PDF)

长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

Issue:
2013年02期
Page:
79-84
Research Field:
交通工程
Publishing date:

Info

Title:
A method of real-time traffic background updating based on block-counting
Author(s):
XUE Ru1; 2;  SONG Huan-sheng1; 3;  FU Yang1
1. School of Information Engineering, Chang'an University, Xi'an 710064, Shaanxi, China; 2. School of Information Engineering, Tibet Institute for Nationalities, Xianyang 712082, Shaanxi, China; 3. Shaanxi Engineering and Technique Research Center for Road and Traffic Detection, Xi'an 710064, Shaanxi, China
Keywords:
traffic engineering;  background updating;  background model;  real-time;  Gaussian mixture model
PACS:
U491.116; TP391
DOI:
-
Abstract:
In order to improve the detecting quality for moving objects, a novel selective background updating method based on block counting was proposed, which could learn the pixel-block change online. First of all, the traditional median filter method was improved, and the initial background of video based on the method was established; then, the difference was achieved by subtraction between corresponding pixel-block of the adjacent frames in the training sequences; for the pixel-blocks whose differences were in threshold value were accumulated and stored in the counter. Through monitoring the counter in real-time, the algorithm could update background according to the value of pixel-block in current frame and background. The results show that compared with the standard codebook, the proposed method's learning speed, foreground extraction rate and accuracy in video 2 are increased by 10%, 3% and 5%, respectively. Compared with GMM, Codebook and S&KB, the speed of the method is more quick than other methods; besides, the method has higher discrimination and its detected error pixel is less,according to test results about the true alarm rate and the error alarm rate. 5 figs, 14 refs.

References:

[1] Vacchetti L,Lepetit V,Fua F.Stable real-time 3D tracking using online and offline information[J].Pattern Analysis and Machine Intelligence,2004,26(10):1385-1391.
[2]Lee P H,Chiu T H,Lin Y L,et al.Real-time pedestrian and vehicle detection in video using 3D cues[C]//IEEE.2009 IEEE International Conference on Multimedia and Exp(ICME).New York:IEEE press,2009:614-617.
[3]Ghosh N,Bhanu B.Incremental unsupervised three-dimensional vehicle model learning from video[J].Intelligent Transportation Systems, 2010,11(2):423-440.
[4]Leotta M J,Mundy J L.Vehicle surveillance with a generic,adaptive,3D vehicle model [J].Pattern Analysis and Machine Intelligence,2011,33(7):1457-1469.
[5]Nicholas A M,Iphigenia K,Chris T K.A background subtraction algorithm for detecting and tracking vehicles[J].Expert Systems with Applications,2011,38(3):1619-1631.
[6]Movshovitz A Y,Peleg S.Bacteria-Filters:persistent particle filters for background subtraction [C]//IEEE.IEEE International Conference Image Processing(ICIP),Hongkong:IEEE press,2010,26-29 Sept:677-680.
[7]Bianco A,Giaccone P,Leonardi E,et al.A Framework for differential frame-based matching algorithms in input-queued switches[C]//IEEE.Twenty-third AnnualJoint Conference of the IEEE Computer and Communications Societies(INFOCOM),Italy:IEEE Press,2004:1147-1157.
[8]Haritaogu I,Harwood D,Davis L.Real-time surveillance of people and their activities[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2000,22(8):809-830.
[9]Wayne Power P,Johann A.Schoonees.Understanding background mixture models for foreground segmentation[C]//IEEE.Proceedings of Image and Vision Computing.Auckland:IEEE press,2002:267-271.
[10]Zivkovic Z.Improved adaptive gaussian mixture model for background subtraction[C]//IEEE.Proceedings of the 17th International Conference on Pattern Recognition.Cambridge:IEEE press,2004,23-26.
[11]Ismail H,David H,Larry S.Real-time surveillance of people and their activities[J].Pattern Analysis and Machine Intelligence,2000,22(8):809-830.
[12]张 丽,李志能.自适应背景更新模型基于在HSV空间阴影检测车辆检测[J].中国图象图形学报,2003,8(7):778-782. ZHANG li,LI Zhi-neng.Adaptive background update model based on shadow detection in HSV space for vehicle detection[J].Journal of Image and Graphics,2008,8(7):778-782.(in Chinese)
[13]Cucchiara R,Grana C,Piccardi M,et al.Statistic and knowledge-based moving object detection in traffic scene[C]//IEEE.Proceedings of the 3th IEEE Conference on Intelligent Transportation Systems.Dearborn:USA IEEE Computer Society Press,2000:27-32.
[14]Kim K,Chalidabhonse T H,Harwood D,et al.Real-time foreground-background segmentation using codebook model[J].Real-Time Imaging,2005,11(3):167-256.

Memo

Memo:
-
Last Update: 2013-04-20