|Table of Contents|

Motion object tracking algorithm using an improved geometric active contour model and Kalman filtering(PDF)

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

Issue:
2011年03期
Page:
90-94
Research Field:
Publishing date:
2011-06-30

Info

Title:
Motion object tracking algorithm using an improved geometric active contour model and Kalman filtering
Author(s):
LI Han1;  WU Qi-sheng2;  LUO Xiang-long1
1. School of Information Engineering, Chang'an University, Xi'an 710064, Shaanxi, China; 2. School of Electronic and Control Engineering, Chang'an University, Xi'an 710064, Shaanxi, China
Keywords:
pattern recognition;  motion object tracking;  geometric active contour;  level set method;  Kalman filtering
PACS:
TP391.41
DOI:
-
Abstract:
In order to overcome the initialization and inaccuracy problem of geometric active contour model in tracking, an improved method is proposed. The statistical feature of inter-frame difference is used to detect automatically the moving area of vehicles in a video sequence. Then this rectangle region of objects is set as the initialization of geometric active contour to fit the edge of objects, which can simplify the initialization. Lastly, an accurate contour of vehicles can be used for Kalman filtering tracking. Taking bounding rectangles as initial curve will simplify the initialization and improve the convergence rate. Through introducing an enforcement item to the level set function, the fitted contour can converge to concave edges, which also increase the convergence performance. The accurate contours can improve the tracking efficiency. 1 tab, 4 figs, 11 refs.

References:

[1] Peterfreund N.Robust tracking of position and velocity with Kalman snakes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(6):564-569.
[2]郭礼华,袁晓彤,李建华.基于直方图的snake视频对象跟踪算法[J].中国图像图形学报,2005,10(2):197-202. GUO Li-hua,YUAN Xiao-tong,LI Jian-hua.Video object tracking method based on snake model using object's histogram information[J].Journal of Image and Graphics,2005,10(2):197-202.
[3]罗 嘉,韦志辉.基于几何轮廓模型的目标跟踪与快速运动估计[J].中国图像图形学报,2009,14(7):1361-1369. LUO Jia,WEI Zhi-hui.Tracking and fast motion estimation via geometric active contour model[J].Journal of Image and Graphics,2009,14(7):1361-1369.
[4]张晓燕,赵荣椿,马志强.基于改进活动轮廓的视频对象自动分割及跟踪算法[J].中国图像图形学报,2007,12(3):438-443. ZHANG Xiao-yan,ZHAO Rong-chun,MA Zhi-qiang.Automatic video segmentation and tracking using a improved active contour[J].Journal of Image and Graphics,2007,12(3):438-443.
[5]Kass M,Witkin A,Terzopolos D.Snake:active contour models[J].International Journal of Vision,1998,1(4):321-331.
[6]Caselles V,Catte F,Coll T,et al.A geometric model for active contours in image processing[J].Numeric Mathematic,1993,66(1):1-31.
[7]Malladi R,Sethian J A,Vemuri B C.Shape modeling with front propagation:a level set approach[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1995,17(2):158-174.
[8]Chan T F,Vese L.Active contours without edges[J].IEEE Transactions on Image Processing,2001,2(10):266-277.
[9]Paragios N,Deriche R.Geodesic active contours and level sets for the detection and tracking of moving objects[J].IEEE Transactions on Pattern Analysis and Mathematics Intelligence,2000,22(3):266-280.
[10]Meuter M,Iurgel U,Park S,et al.The un-scented Kalman filter for pedestrian tracking from a moving host[J].IEEE Intelligent Vehicles Symposium,2008(6):37-42.
[11]杨艳芳,齐美彬,王 倩,等.基于运动检测与运动搜索的多目标跟踪[J].计算机工程,2008,34(19):222-224. YANG Yan-fang,IQ Mei-bin,WANG Qian,et al.Multi-object tracking based on motion detection and motion search[J].Computer Engineering,2008,34(19):222-224.

Memo

Memo:
-
Last Update: 2011-06-30