A statistical approach for object motion estimation with MPEG motion vectors

被引：0

作者：

Yu, XD ^{[1
]}

Xue, P ^{[1
]}

Tian, Q ^{[1
]}

机构：

[1] Nanyang Technol Univ, Sch Elect & Elect Engn, Singapore 2263, Singapore

来源：

2004 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXP (ICME), VOLS 1-3 | 2004年

关键词：

D O I：

10.1109/ICME.2004.1394243

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper we propose a statistical approach to estimate the object motion with MPEG motion vectors. A model with two normal distribution terms is applied to represent the simplified object motion. One term models the noises embedded in the motion vector field produced in the encoding stage and the other term models the randomness of the trite object motion. Experiments with vehicle motion estimation from MPEG traffic video are used to evaluate the proposed algorithm. The influence of time window, frame size and reference frame distance are investigated. The vehicle speeds can be estimated with a high accuracy tip to 85% similar to 92%.

引用

页码：519 / 522

页数：4

共 10 条

[1] Babu RV, 2002, INT CONF ACOUST SPEE, P3788
[2] BARTOLINI F, 1996, P IEEE INT C IM PROC, P87
[3] MOTION RECOVERY FOR VIDEO CONTENT CLASSIFICATION
DIMITROVA, N
GOLSHANI, F
[J]. ACM TRANSACTIONS ON INFORMATION SYSTEMS, 1995, 13 (04) : 408 - 439
[4] ENG HL, 1999, IEEE T CIRCUIT SYSTE, P284
[5] Optimal projection of 2-D displacements for 3-D translational motion estimation
Garcia, C
Tziritas, G
[J]. IMAGE AND VISION COMPUTING, 2002, 20 (11) : 793 - 804
[6] GONZALES CA, 1999, IBM J RES DEV, V43
[7] WANG J, 2000, IS T SPIE S EL IM VI
[8] YOON K, 2000, INT C PATT RECOG, V1, P1819
[9] Yu XD, 2002, IEEE 5TH INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS, PROCEEDINGS, P37, DOI 10.1109/ITSC.2002.1041185
[10] ZEN H, 1999, P IEEE INT C IM PROC, V4, P25

← 1 →