High-velocity optical flow

被引:0
作者
Vergeest, Joris [1 ]
机构
[1] Delft Univ Technol, Landbergstr 15, Delft, Netherlands
来源
WSCG 2013, COMMUNICATION PAPERS PROCEEDINGS | 2013年
关键词
Optical flow; high speed; noise; low resolution;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Optical flow is widely used to estimate the velocity of objects relative to a digital camera. Most commonly, two images taken with the same camera at small time difference are compared in order to detect the displacement of structures in 2D image space. Such displacement could be a measure of displacement, or motion, of objects in the scene relative to the camera. At high velocities, the displacement in image space is relatively large and the correlation of image structures gets more difficult. The displacement can be reduced by reducing the time difference, or increasing the number of frames taken per second. However, due to the reduced exposure time, the quality of the individual images gets poorer. In some practical situations, it appears technically very difficult to achieve reliable speed measurement at high velocities, even when using high-speed cameras. One example is the measurement of self-speed from images of the road surface taken with a camera from a driving car. In view of this purpose we explore the potential of the method and its limitations.
引用
收藏
页码:112 / 119
页数:8
相关论文
共 10 条
[1]  
Barron J.L., 2013, INT J COMPUT VISION, V12, P43
[2]  
Mammarella M., IEEE T SYSTEMS MAN C, V42, P1752
[3]  
Nieto M., 2010, MACHINE VISION APPL
[4]  
Schakel W.J., P 13 INT IEEE ANN C, P759
[5]  
Souhila K., 2007, International Journal of Advanced Robotic Systems, V4, P13
[6]   Memory effects in microscopic traffic models and wide scattering in flow-density data [J].
Treiber, M ;
Helbing, D .
PHYSICAL REVIEW E, 2003, 68 (04)
[7]  
Vergeest J.S.M., P WSCG 2012 C
[8]  
Vergeest J.S.M., 2012 IEEE INT VERH S, P45
[9]  
Work D.B., 2010, TRANSPORT RES C-EMER, P568
[10]   Use Patterns Among Early Adopters of Adaptive Cruise Control [J].
Xiong, Huimin ;
Boyle, Linda Ng ;
Moeckli, Jane ;
Dow, Benjamin R. ;
Brown, Timothy L. .
HUMAN FACTORS, 2012, 54 (05) :722-733